MiwmmmMwmamamwmi
dtwttr.
Page
Quality Profile \ Clarke American Checks, Inc.
5
Motorola, Inc.
5
Chugach School District
51
SSM Health Care
51
Texas Nameplate, Inc.
93
Sundaram-Clayton
93
Custom Research Incorporated
153
Bl
153
Solar Turbines, Inc.
205
Corning Telecommunications Products Division
205
Sunny Fresh Foods
255
Merrill Lynch Credit Corporation
255
STMicroelectronics, Inc.-Region Americas
315
Boeing Aircraft and Tanker Programs
315
Wainwright Industries, Inc.
373
ADAC Laboratories
373
American Electric Power
433
Pal's Sudden Service
433
Karlee Company
481
Los Alamos National Bank
481
Granite Rock Company
517
Branch-Smith Printing Division
517
Dana Corporation-Spicer Driveshaft
567
3M Dental Products Division
567
Armstrong World Industries Building Products Operations
637
Xerox Business Services
637
Trident Precision Manufacturing, Inc.
689
Operations Management International, Inc.
689
(X. Businesses and Organizations Cited in this Book
3M Dental Products Division ADAC Laboratories Allied Signal Amazon.com American Electric Power American Express American National Standards Institute American Parkinson's Disease Association American Red Cross American Society for Quality Ames Rubber Corporation Analog Devices, Inc. Apple Computer Armstrong Building Products Operations AT&T Australian Quality Council Avis Baxter Healthcare International Bell System HE
Big Bear Stores Bloomfield Tool Company Black & Decker Boeing Airlift and Tanker Boise Cascade Bose Corporation Branch-Smith Printing Division CapStar Health Systems Cargill, Inc. Caterpillar, Inc. Center for Quality of Management Chase Manhattan Bank Chemical Workers Association Child Focus, Inc. Cincinnati Water Works Clarke American Checks CNH Capital Coca-Cola Company Continental Airlines Convergys Corporation Coors Brewing Company Copeland Companies Corryville Foundry Company CRI Star Crystal Silicon Custom Research Inc. Daimler-Chrysler Dana Corporation Deer Valley Resort Defense Supply Center Philadelphia
Dell Computer Disney Corporation Domino's Pizza Douglas Aircraft
fat
^ XN’ , ,
Eastman Chemical Company European Foundation for Quality Management Excelsior Inn Fanuc Ltd. Federal Quality Institute FedEx Fidelity Investments First National Bank of Chicago Florida Power and Light Ford Motor Company Federal Quality Institute Froedtert Memorial Lutheran Hospital General Electric The Gap General Motors Powertrain Division GeoOrb Polymers, North America Gold Star Chili, Inc. Granite Rock Company GTE Directories Corporation Herend Porcelain Manufacturing Hershey Foods Corporation Hewlett Packard Hillerich & Bradby Co. Hilton Hotels Honeywell, Inc. Honeywell International Hydraulic Lift Company Hyundai Motor Co. IBM Rochester Ina Tile Company Institute for Healthcare Improvement Internal Revenue Service International Organization for Standardization ITT Janson Medical Clinic Joint Commission on Accreditation of Healthcare Organizations Jim's SteakHouse Johnson & Johnson Johnson Controls, Inc. Juran Institute continued on next page
Businesses and Organizations Cited in this Book (continued) JUSE (Union of Japanese Scientists and Engineers) KARLEE Company Kelly's Seafood Restaurant Koalaty Kid La Ventana Window Company LaRosa's, Inc. Legal Sea Foods L.L. Bean Los Alamos National Bank Lucas Suminomo Brakes, Inc. Magnivision Marlow Industries Master Black Belts McDonald's McDonnell Douglas Merrill Lynch Credit Corporation Microsoft Corporation Middletown Regional Hospital Midwest Express Airlines Mitsubishi Motorola, Inc. Murphy Trucking, Inc. National Cash Register Company National Committee for Quality Assurance National Furniture National Institute of Standards and Technology National Labor Relations Board National Quality Institute (Canada) National Quality Program Nationwide Insurance NCR Corporation Nissan Nordstrom Nucor Corporation Operational Management International, Inc. Palmer Sausage Co. Pal's Sudden Service Penn State University PepsiCo PIMS Associates Polaroid Procter & Gamble Prudential Insurance Company Rath & Strong Raytheon
Readilunch Restaurant Red Cross The Ritz-Carlton Hotel Company Rotor Clip Company, Inc. Royal Mail (UK) Rubbermaid Samsung Electronics Co. SAS Institute, Inc. Selit Corp. Semco S/A Shure, Inc. Siemens Energy and Automation Solar Turbines, Inc. Solectron Corporation Southwest Airlines Southwest Louisiana Regional Medical Center Starbucks STMicroelectronics Stuart Injection Molding Co. Sun Microsystems Sundaram-Clayton Sunny Fresh Foods Sunset Manufacturing, Inc. Superquinn TD Industries Techneglas TecSmart Electronics Texas Instruments (TI) Texas Nameplate Company Torque Traction Technologies, Inc. Toyota Motor Corporation Trident Precision Manufacturing, Inc. TVS Partnership Proprietary, LTD Ultra-Productivity Fasteners Company Unison Industries, Inc. W. Edward Deming Institute Wainwright Industries Wal-Mart Walt Disney Company Westel Mobile Telecommunications Co. Ltd. Westerfield Construction Western Electric Company Whirlpool Wilson Sporting Goods Xerox Corporation Business Services For more information, go to http: //www.evans.swlearning.com
The Management and Control of Quality Sixth Edition
James R. Evans University of Cincinnati
William M. Lindsay Northern Kentucky University
THOMSON
-*--SOUTH-WESTERN
Australia
Canada
Mexico
■
Singapore
Spain
United
Kingdom
United States
3 HOMSON —--
SOUTH-WESTERN
The Management and Control of Quality, 6e
James R. Evans and William M. Lindsay VP/Editorial Director:
Production Editor:
Production House:
Jack W. Calhoun
Chris Sears
Trejo Production
VP/Editor-in-Chief:
Manufacturing Coordinator:
Cover Designer:
George Werthman
Diane Lohman
Bethany Casey
Acquisitions Editor:
Technology Project Editor:
Cover Image:
Charles E. McCormick, Jr.
Christine A. Wittmer
PhotoDisc, Inc.
Developmental Editor:
Media Editor:
Internal Designer:
Alice C. Denny
Amy Wilson
Jennifer Lambert, Jen2 Design
Marketing Manager:
Design Project Manager:
Printer:
Larry Qualls
Bethany Casey
Thomson-West Eagan, MN
COPYRIGHT © 2005 by South-Western, part of the Thomson Corporation. SouthWestern, Thomson, and the Thomson logo are trademarks used herein under license.
ALL RIGHTS RESERVED.
For permission to use material from this text or product, submit a request online at http: / / www.thomsonrights.com. Any additional questions about per¬ missions can be submitted by email to [emailprotected].
Printed in the United States of America 1 2 3 4 5 07 06 05 04 Package ISBN: 0-324-20223-7 Book Only ISBN: 0-324-20224-5 Student CD ISBN: 0-324-20225-3 Infotrac ISBN: 0-534-55853-5 Library of Congress Control Number: 2003113145
No part of this work covered by the copyright hereon may be repro¬ duced or used in any form or by any means—graphic, electronic, or mechanical, including photo¬ copying, recording, taping, Web distribution or information storage and retrieval systems—without the written permission of the publisher.
For more information contact South-Western, 5191 Natorp Boulevard, Mason, Ohio, 45040. Or you can visit our Internet site at: http:/ / www.swlearning.com
Brief Contents Preface
xvii
Q The Quality System
/
Chapter 1
Introduction to Quality
V
Chapter 2
Total Quality in Organizations
49
1/ Chapter 3
Philosophies and Frameworks
91
Ll
3
149
The Management System
-1
Chapter 4
Focusing on Customers
Chapter 5
Leadership and Strategic Planning
Chapter 6
Human Resource Practices
Chapter 7
Process Management
Chapter 8
Performance Measurement and Strategic Information Management
Chapter 9
Building and Sustaining Total Quality Organizations
3
151 203
253
313
477
Six Sigma and the Technical System
Chapter 10
Principles of Six Sigma
Chapter 11
Statistical Thinking and Applications
Chapter 12
Design for Six Sigma
Chapter 13
Tools for Process Improvement
Chapter 14
Statistical Process Control
'§!;[ Appendixes
371
431
479 515
565 635
687
A-1
Digitized by the Internet Archive in 2018 with funding from Kahle/Austin Foundation
https://archive.org/details/managementcontroOOjame
Contents
Q The Quality System
Chapter 1 Introduction to Quality 3 The History and Importance of Quality 4 The Age of Craftsmanship 4
Quality Profiles: Clarke American Checks, Inc., and Motorola, Inc. 5 The Early Twentieth Century 6 Post-World War II 7 The U.S. "Quality Revolution" 8 Early Successes 9 From Product Quality to Performance Excellence 10 Disappointments and Criticism 10 Current and Future Challenges
11
Defining Quality 12 Judgmental Perspective 12 Product-Based Perspective 13 User-Based Perspective 13 Value-Based Perspective 13 Manufacturing-Based Perspective 14 Integrating Perspectives on Quality 14 Customer-Driven Quality 16
Quality as a Management Framework 17 Principles of Total Quality 18 Infrastructure, Practices, and Tools 23
Quality and Competitive Advantage 25 Quality and Business Results 27
Three Levels of Quality 29 Quality and Personal Values 29
Quality in Practice: The Evolution of Quality at Xerox: From Leadership Through Quality to Lean Six Sigma
32
Quality in Practice: Bringing Total Quality Principles to Life at
KARLEE
37
Review Questions 40 Discussion Questions 40 Projects, Etc. 41 CASES Skilled Care Pharmacy 43 A Tale of Two Restaurants 44 A Total Quality Business Model 45 Endnotes 47 Bibliography 48
Chapter 2 Total Quality in Organizations 49 Quality and Systems Thinking 50
Quality Profiles: Chugach School District and SSM Health Care
51
Quality in Manufacturing 52
Manufacturing Systems 53 Quality in Services 58
Contrasts with Manufacturing 59 Components of Service System Quality 60 Quality in Health Care 63 Quality in Education 65
Koalaty Kid 67 Quality in Higher Education 69 Quality in Small Businesses and Not-for-Profits 72 Quality in the Public Sector 74
Quality in the Federal Government 74 State and Local Quality Efforts 76 Quality in Practice: Service Quality at The Ritz-Carlton Hotel Company Quality in Practice: Pearl River School District
80
Review Questions 82 Discussion Questions 83 Projects, Etc. 84 CASES Toyota Motor Corporation, Ltd. 85 The Nightmare on Telecom Street 87 Child Focus, Inc. 87 Endnotes 88 Bibliography 90
Chapter 3 Philosophies and Frameworks 91 The Deming Philosophy 92
Quality Profiles: Texas Nameplate, Inc., and Sundaram-Clayton Foundations of the Deming Philosophy 94
93
78
Deming's 14 Points 100 The Juran Philosophy 106 The Crosby Philosophy 108 Comparisons of Quality Philosophies 110 Other Quality Philosophers 110
A. V. Feigenbaum 110 Kaoru Ishikawa 111 Genichi Taguchi 112 Quality Management Awards and Frameworks 113 The Malcolm Baldrige National Quality Award 113
History and Purpose 114 The Criteria for Performance Excellence 114 Criteria Evolution 120 Using the Baldrige Criteria 121 Impacts of the Baldrige Program 122 Baldrige and Deming 123 International Quality Award Programs 123
The Deming Prize 123 European Quality Award 124 Canadian Awards for Business Excellence 126 Australian Business Excellence Award 126 ISO 9000:2000 128 Structure of the ISO 9000:2000 Standards 128 Factors Leading to ISO 9000:2000 130 Implementation and Registration 131 Benefits of ISO 9000 132 Six Sigma 132
Evolution of Six Sigma 132 Six Sigma as a Quality Framework 134 Baldrige, ISO 9000, and Six Sigma 135
Quality in Practice: Building Business Excellence in Hungary Quality in Practice: Six Sigma Integration at Samsung
137
139
Review Questions 140 Discussion Questions 141 Projects, Etc. 142 CASES TecSmart Electronics 144 Can Six Sigma Work in Health Care? 145 CapStar Health System, Inc.: Understanding the Organizational Environment 146 GeoOrb Polymers, North America: Understanding the Organizational Environment 146 Endnotes 146 Bibliography 148
2
The Management System
_149_
Chapter 4 Focusing on Customers 151 The Importance of Customer Satisfaction and Loyalty 152
Quality Profiles: Custom Research Incorporated and
Bl
153
The American Customer Satisfaction Index 155 Creating Satisfied Customers 156
Leading Practices 158 Identifying Customers 160
Customer Segmentation 161 Understanding Customer Needs 162 Gathering and Analyzing Customer Information 166 Customer Relationship Management 168
Accessibility and Commitments 169 Selecting and Developing Customer Contact Employees 169 Customer Contact Requirements 171 Effective Complaint Management 171 Strategic Partnerships and Alliances 173 Exploiting CRM Technology 174 Measuring Customer Satisfaction 175
Designing Satisfaction Surveys 175 Analyzing and Using Customer Feedback 179 Why Many Customer Satisfaction Efforts Fail 181 Customer Perceived Value 182 Customer Focus in the Baldrige Criteria,
ISO 9000,
and Six Sigma
Quality in Practice: Understanding the Voice of the Customer at LaRosa’s Pizzerias Quality in Practice: Customer Focus at Amazon.com
186
Review Questions 188 Discussion Questions 189 Projects, Etc. 192 CASES The Case of the Missing Reservation 194 American Parkinson's Disease Association Center 194 Gold Star Chili: Customer and Market Knowledge 195 CapStar Health Systems: Customer Focus 199 Endnotes 199 Bibliography 201
Chapter 5
Leadership and Strategic Planning 203
Leadership for Quality 204
Quality Profiles: Solar Turbines, Inc., and Corning Telecommunications Products Division
205
Leading Practices for Leadership 208 Leadership Theory and Practice 211
viii
183 184
Contemporary and Emerging Leadership Theories 212 Applying Leadership Theory in a TQ Environment 213 Creating the Leadership System 214
Leadership and Social Responsibilities 218 Strategic Planning 219
Leading Practices for Strategic Planning 219 Strategy Development 222 Strategy Deployment 224 Linking Human Resource Plans and Business Strategy 227 The Seven Management and Planning Tools 228 Leadership, Strategy, and Organizational Structure 229 Leadership and Strategic Planning in the Baldrige Criteria, and
Six
ISO
Sigma 232
Quality in Practice: Leadership in the Virgin Group
234
Quality in Practice: Strategic Planning at Branch-Smith Printing Division
237
Review Questions 240 Discussion Questions 240 Problems 242 Projects, Etc. 245 CASES Johnsonville Foods 246 A Strategic Bottleneck 247 Corryville Foundry Company 248 CapStar Health Systems: Leadership and Strategic Planning 249 Endnotes 249 Bibliography 251
Chapter 6 Human Resource Practices 253 Quality Profiles: Sunny Fresh Foods and Merrill Lynch Credit Corporation
255
The Scope of Human Resource Management 256
Leading Practices 257 Teams in Organizational Design and Quality Improvement 262
Building Effective Teams 266 Six Sigma Project Teams 268 Designing High-Performance Work Systems 269
Work and Job Design 273 Employee Involvement 275 Empowerment 278 Recruitment and Career Development 280 Training and Education 281 Compensation and Recognition 282 Health, Safety, and Employee Well-Being 286 Motivating Employees 287 Performance Appraisal 289
9000,
Measuring Employee Satisfaction and HRM Effectiveness 291 HRM in the Internet Age 292 Human Resource Focus in the Baldrige Criteria, Sigma
ISO 9000,
and Six
294
Quality in Practice: Quality in Practice:
TD Industries 295 L.L. Bean 297
Review Questions 299 Discussion Questions 300 Projects, Etc. 303 CASES The Hopeful Telecommuter 304 Crystal Silicon, Inc. 305 TVS Partnership Proprietary, Ltd., Brisbane, Australia 306 CapStar Health Systems: Human Resource Focus 308 Endnotes 308 Bibliography 311 Chapter
7
Process Management
313
Quality Profiles: STMicroelectronics, Inc—Region Americas, and Boeing Aircraft and Tanker Programs 315 The Scope of Process Management 316
Leading Practices 318 Product Design Processes 323
Cost, Manufacturability, and Quality 325 Design Quality and Social Responsibility 326 Streamlining the Product Development Process 329 Designing Processes for Quality 331
Special Considerations in Service Process Design. 332 Projects as Value-Creation Processes 335
Project Life Cycle Management 336 Process Control 339
Process Control in Services 342 Process Improvement 345
Kaizen 347 Flexibility and Cycle Time Reduction 348 Breakthrough Improvement 350 Process Management in the Baldrige Criteria,
ISO 9000,
Six Sigma 352
Quality in Practice: Gold Star Chili: Process Management 354 Quality in Practice: Bringing Process Management to Education Review Questions 359 Discussion Questions 360 Projects, Etc. 362 CASES The State University Experience 363
x
357
and
The PIVOT Initiative at Midwest Bank, Part I 364 Stuart Injection Molding Company 366 CapStar Health Systems: Process Management 367 Endnotes 367 Bibliography 369
Chapter 8 Performance Measurement and Strategic Information Management 371 Quality Profiles: Wainwright Industries, Inc. and
ADAC
Laboratories
373
The Strategic Value of Information 374
Leading Practices 375 The Scope of Performance Measurement 378
The Balanced Scorecard 379 Customer-Focused Measures 383 Product and Service Measures 383 Financial and Market Measures 383 Human Resource Measures 384 Organizational Effectiveness Measures 384 Governance and Social Responsibility Measures 385 The Role of Comparative Data 385 Designing Effective Performance Measurement Systems 386
Linking Measures to Strategy 387 Process-Level Measurements 390 Identifying and Selecting Process Measures 391 Aligning Strategic and Process-Level Measurements 393 Analyzing and Using Performance Data 395 The Cost of Quality 398
Quality Cost Classification 398 Quality Costs in Service Organizations 401 Capturing Quality Costs Through Activity-Based Costing 401 Measuring the Return on Quality 402 Managing Information and Knowledge Assets 403
Data Validity 403 Data Accessibility and Security 404 Knowledge Management 405 Measurement and Information Management in the Baldrige Criteria, ISO 9000, and Six Sigma 408
Quality in Practice: Knowledge Management for Continuous Improvement AT CONVERGYS
410
Quality in Practice: Modeling Cause-and-Effect Relationships at IBM Rochester Review Questions 416 Discussion Questions 416 Problems 417
414
Projects, Etc. 423 CASES Coyote Community College 423 Ultra-Productivity Fasteners, Part I 427 CapStar Health Systems: Information and Analysis 427 Endnotes 427 Bibliography 429
Chapter 9 Building and Sustaining Total Quality Organizations 431
TQ 432
Making the Commitment to
Quality Profiles: American Electric Power and Pal’s Sudden Service
433
Organizational Culture and Total Quality 434
Cultural Change 438 Building on Best Practices 439 The Role of Employees in Cultural Change 441 Change Management 444
Implementation Barriers to Creating a TQ Culture 445 Sustaining the Quality Organization 448
Quality as a Journey 448 The Learning Organization 450 Self-Assessment Processes 454
Leveraging Self-Assessment: The Importance of Follow-Up 458 Implementing ISO
A
9000,
Baldrige, and Six Sigma
View Toward the Future
459
463
Quality in Practice: The Eastman Way
464
Quality in Practice: Merging Divergent Quality Systems at Honeywell Review Questions 469 Discussion Questions 470 Projects, Etc. 471 CASES The Parable of the Green Lawn 472 The Yellow Brick Road to Quality 473 Westerfield Construction 473 Endnotes 474 Bibliography 476
Six
Sigma and the Technical System Chapter 10 Principles of Six Sigma 479 The Statistical Basis of Six Sigma
480
Quality Profiles: Karlee Company and Los Alamos National Bank Project Selection for Six Sigma
Six
Sigma Problem Solving
486
The DMAIC Methodology 488
484
481
466
Tools and Techniques 492 Design for Six Sigma 492 Team Processes and Project Management 494 Six Sigma in Services and Small Organizations 494 Six Sigma and Lean Production 496 Lean Six Sigma and Services 498
Quality in Practice: An Application of Six Sigma to Reduce Medical Errors Quality in Practice: Ford’s Drive to Six Sigma Quality 502
500
Review Questions 503 Discussion Questions 504 Problems 505 Projects, Etc. 506 Cases Implementing Six Sigma at GE Fanuc 506 The PIVOT Initiative at Midwest Bank, Part 2 508 Endnotes 512 Bibliography 513
Chapter 11 Statistical Thinking and Applications 515 Statistical Thinking 516
Quality Profiles: Granite Rock Company and Branch-Smith Printing Division Deming's Red Bead and Funnel Experiments 520 Statistical Foundations 526
Random Variables and Probability Distributions 526 Sampling 527 Statistical Methodology 531
Descriptive Statistics 532 Statistical Analysis with Microsoft Excel 533 Statistical Inference 537 Enumerative and Analytic Studies 542 Design of Experiments 542 Analysis of Variance (ANOVA) 546 Regression and Correlation 548
Quality in Practice: Improving Quality of a Wave Soldering Process Through Design of Experiments 550 Quality in Practice: Applying Statistical Analysis in a Six Sigma Project at
GE
Fanuc
552
Review Questions 555 Problems 556 Projects, Etc. 559 CASES The Disciplinary Citation 560 The Quarterly Sales Report 560 The HMO Pharmacy Crisis 562 Endnotes 563 Bibliography 564
517
Chapter 12 Design for Six Sigma 565 Tools for Concept Development 566
Quality Profiles: Dana Corporation-Spicer Driveshaft, and 3M Dental Products Division
567
Quality Function Deployment 568 Concept Engineering 578 Tools for Design Development 580
Design Failure Mode and Effects Analysis 582 Reliability Prediction 582 Tools for Design Optimization 594
The Taguchi Loss Function 594 Optimizing Reliability 597 Tools for Design Verification 598
Reliability Testing 598 Measurement System Evaluation 599 Process Capability Evaluation 606 Quality in Practice: Testing Audio Components at Shure, Inc.
613
Quality in Practice: Applying Quality Function Deployment to a University Support Service
615
Review Questions 619 Problems 620 Projects, Etc. 627 CASES Hydraulic Lift Co. 629 Bloomfield Tool Co. 632 Endnotes 633 Bibliography 634
Chapter 13 Tools for Process Improvement 635 Process Improvement Methodologies 636
The Deming Cycle 636 Quality Profiles: Armstrong World Industries Building Products Operations and Xerox Business Services
637
FADE 640 Juran's Breakthrough Sequence 640 Creative Problem Solving 641 Basic Tools for Process Improvement 641
Flowcharts 642 Run Charts and Control Charts 644 Check Sheets 648 Histograms 649 Pareto Diagrams 651 Cause-and-Effect Diagrams 654 Scatter Diagrams 654
Other Tools for Process Improvement 658
Kaizen Blitz 658 Poka-Yoke (Mistake-Proofing) 658 Process Simulation 661 Engaging the Workforce in Process Improvement 663
Skills for Team Leaders 665 Skills for Team Members 666
Quality in Practice: Process Improvement on the Free-Throw Line
667
Quality in Practice: Improving Patient Services at Middletown Regional Hospital
669
Review Questions 671 Discussion Questions 671 Problems 676 Projects, Etc. 680 CASES Readilunch Restaurant 681 National Furniture 682 Janson Medical Clinic 682 Endnotes 684 Bibliography 685
Chapter 14 Statistical Process Control 687 Quality Profiles: Trident Precision Manufacturing, Inc. and Operations Management International, Inc. 689 Quality Control Measurements 690 Capability and Control 690 SPC Methodology 693 Control Charts for Variables Data 694
Constructing x and R-Charts and Establishing Statistical Control 694 Interpreting Patterns in Control Charts 695 Process Monitoring and Control 708 Estimating Process Capability 710 Modified Control Limits 712 Excel Spreadsheet Templates 712 Special Control Charts for Variables Data 713
x and s-Charts 713 Charts for Individuals 716 Control Charts for Attributes 721
Fraction Nonconforming (p) Chart 721 Variable Sample Size 723 Mp-Charts for Number Nonconforming 727 Charts for Defects 729 Choosing Between c- and w-Charts 732 Summary of Control Chart Construction 734 Designing Control Charts 735
XV
Basis for Sampling 735 Sample Size 736 Sampling Frequency 737 Location of Control Limits 737 SPC, ISO 9000:2000, and Six Sigma 739 Controlling Six Sigma Processes 740 Pre-Control 740
Quality in Practice: Applying SPC to Pharmaceutical Product Manufacturing 742 Quality in Practice: Using a //-Chart in a Receiving Process 746 Review Questions 749 Problems 750 CASES La Ventana Window Company 757 Murphy Trucking, Inc. 758 Day Industries 759 Endnotes 759 Bibliography 760 Appendixes A
Areas for the Standard Normal Distribution A-1
B
Factors for Control Charts A-3
C
Random Digits A-4
D
Binomial Probabilities A-5
E
Poisson Probabilities A-10
F
Values of
e~m
A-15
Solutions to Even-Numbered Problems S-1 Index 1-1
Preface
"Why is quality still so bad?" laments Scott Paton, editor-in-chief of Quality Digest, a major trade publication for the quality profession, in his April 2002 editorial. Although he notes that the quality of U.S. products as a whole is better now than it was in 1972, it is worse than it was in 1992 (when quality was the buzzword among businesses). And it's not just in manufacturing. Paton states, "If you've had a truly high-quality experience on a recent flight or with your loan application or buying a car or with your hospital, you're in the minority." He places the blame squarely on senior management, who fail to see the simple but essential relationship between customers' needs and expectations and designing, building, and delivering great products and services. We agree completely—the war for better quality must continue—which is why we continue to update and improve this book. Today's business and not-for-profit organizations need to capitalize on the knowledge and "lessons learned" that excel¬ lent organizations have acquired. One of the best ways of obtaining such knowledge is from the national role models that have emerged from the Malcolm Baldrige National Quality Award in the United States and similar programs throughout the world. The expansion of Baldrige to nonprofit education and health care—along with recent winners—has generated a high level of interest among these sectors. Thus in this new edition, we continue to use Baldrige as the fundamental framework for organizing and presenting key issues of performance excellence. Six Sigma* has taken the corporate world by storm and represents the thrust of many efforts to improve products, services, and processes. Moreover, Six Sigma is grounded in the fundamental principles of total quality that have defined the concept for several decades. Hence, the most significant revision to this edition of the book is a comprehensive focus on Six Sigma and its relationships with fundamental quality principles and the Baldrige categories.
CHANGES IN THE SIXTH EDITION The sixth edition of The Management and Control of Quality continues to embrace the fundamental principles and historical foundations of total quality and to promote * Six Sigma is a federally registered trademark and service mark of Motorola, Lac.
Preface
high-performance management practices that are reflected in the Baldrige Criteria, while providing a foundation for understanding and applying Six Sigma. The signif¬ icant changes for the new edition include: • Revised, integrated, and more comprehensive coverage of Six Sigma philos¬ ophy, concepts, and techniques, and new chapters on Principles of Six Sigma and Design for Six Sigma. • Contrasts and comparisons of Baldrige, ISO 9000, and Six Sigma in the man¬ agerial chapters of the book. • New internal layout highlighting important concepts to improve readability. • A new "Bonus Materials" folder on the CD-ROM that includes additional cases, summaries of key points and terminology, and supplementary topics for each chapter. • Text coverage of most of the body of knowledge (BOK) required for ASQ certi¬ fication as a Certified Quality Manager. As in the previous edition, Part 1 introduces fundamentals, and Part 2 concen¬ trates on the management system. However, Part 3 has been refocused around Six Sigma and basic technical topics. Tire chapters in this section have been revised sig¬ nificantly, and new chapters on Six Sigma Principles and Design for Six Sigma have been added. This organization provides the instructor with considerable flexibility in focusing on both managerial and technical topics, for audiences ranging from under¬ graduate students, MBA students, or executives. We have updated every chapter to reflect current thinking and practice, including many new and interesting Quality in Practice cases. Cases and a wide variety of exam¬ ples from organizations around the world emphasize the importance of quality in the global economy. New Organization and Focus on Six Sigma Part 1 provides an introduction to quality management principles. • Chapter 1 introduces the notion of quality, its history and importance, defini¬ tions, basic principles, and its impact on competitive advantage and financial return. • Chapter 2 explores the role of total quality in all key economic sectors: manu¬ facturing, service, health care, education, and the public sector. • Chapter 3 presents the philosophical perspectives supporting total quality, chiefly those of Deming, Juran, and Crosby, as well as quality management frameworks defined by the Malcolm Baldrige National Quality Award and the Criteria for Performance Excellence, ISO 9000, and Six Sigma. Part 2 focuses on the management system, which is concerned with planning to meet customers' needs, arranging to meet those needs through leadership and strategic planning, and accomplishing goals through the actions of people and work processes. All of these activities are done with an eye toward continuous improve¬ ment and using data and information to guide the decision-making process. Each of Chapters 4 through 8 summarizes the key relationships and importance of the topics to Baldrige, Six Sigma, and ISO 9000. • In Chapter 4, the focus is on understanding customers and their needs, and practices to achieve customer satisfaction. • Chapter 5 covers the important role of total quality in leadership and strategic planning.
Preface
• Chapter 6 deals with human resource practices, specifically the design of high-performance work systems and HR management in a total quality environment. Chapter 7 outlines the scope of process management activities for value creation and support processes, the philosophy of continuous improvement, and the role of project management in Six Sigma and other quality improvement efforts. In Chapter 8, the focus is on the use of data and information to measure and manage organizational performance. This chapter includes discussion of bal¬ anced scorecards and recent approaches to knowledge management. • The final chapter in this part. Chapter 9, deals with building and sustaining quality organizations. Coverage includes building a quality infrastructure, organizational culture, and new sections on self-assessment and change management. Part 3 includes basic technical issues, tools, and techniques that underpin Six Sigma. • Chapter 10 is new and provides an overview of the principles of Six Sigma, its problem-solving orientation and methodology, and synergy with lean production. • Chapter 11 provides a general introduction to statistical thinking and the role of statistical tools and methodology in quality and Six Sigma. • Chapter 12 is also new and addresses tools that support the concept of design for Six Sigma (DFSS), including concept development, design development, design optimization, and design verification. • Chapter 13 focuses on quality improvement, including process improvement paradigms, basic tools, and new material on engaging the workforce. • Finally, Chapter 14 provides a comprehensive introduction to statistical process control. Features and Pedagogy to Enhance Learning The sixth edition of The Management and Control of Quality contains a new internal design to make it more "reader-friendly." Each chapter begins with a single-page Quality Profile of two role-model organizations. Significant points of learning and emphasis are now highlighted in distinctive boxes. Quality Spotlight icons in the margin identify examples of specific organizational actions, and CD icons in the margin indicate that extensive supplementary materials may be found on the accom¬ panying Student CD-ROM. The Quality Profile presented at the beginning of each chapter provides back¬ ground, important practices, and results for organizations that embrace total quality principles. Most of these organizations are Baldrige winners. In each chapter. Quality in Practice case studies describe real applications of the chapter material. They rein¬ force the chapter concepts and provide opportunities for discussion and more prac¬ tical understanding. Many of the case studies are drawn from real, published, or personal experiences of the authors. End-of-chapter materials for each chapter include Review Questions, which are designed to help students check their understanding of the key concepts presented in the chapter. All the chapters in Parts 1 and 2 also have Discussion Questions that are open-ended or experiential in nature, and designed to help students expand their thinking or tie practical experiences to abstract concepts. Chapter 8 and those in Part 3 include Problems designed to help students develop and practice quantitative skills.
xix
Preface
XX
Most chapters have a section entitled Projects, Etc. that provides projects in¬ volving field investigation or other types of research. Finally, each chapter includes several Cases, which encourage critical thinking to apply the concepts to unstruc¬ tured or more comprehensive situations.
Student CD-ROM The CD-ROM that comes with new copies of the text contains extensive Bonus Mate¬ rials, including the following: • • • • • • • ® •
Summaries of key points and terminology for each chapter Additional readings that support and extend the presentations in the chapters Baldrige criteria and case studies, Multimedia cases (indicated with an icon in the textbook) with digital videos Additional cases for instruction and discussion Microsoft® Excel templates for quantitative analysis Web links to key organizations cited in the book A Glossary of terms from the textbook The Quality Gamebox™, developed by PQ Systems in Dayton, Ohio, which is a collection of simulations for teaching concepts of variability
InfoTrac® College Edition InfoTrac provides students with complete online access to full-text articles from thou¬ sands of scholarly and popular periodicals. Research possibilities are unlimited. New copies of the sixth edition textbook include a passcode that provides unlimited access for four months.
Web Site The URT for the Web site for new edition of The Management and Control of Quality is http://www.evans.swlearning.com. Among other resources on the site, the Web links for organizations mentioned in the text are given in the Internet Resources. The lists are organized by chapter for convenience and may be used to access the information about the organization. A listing of general Web links on quality is also included.
Note on Company References and Citations In today's ever-changing business environment, many companies and divisions are sold, merged, divested, or declared bankruptcy, resulting in name changes. For example, Texas Instruments Defense Systems & Electronics Group was sold to Raytheon and is now part of Raytheon Systems Company, and AT&T Universal Card Services was bought by CitiBank (which is now CitiGroup). Although we have made efforts to note these changes in the book, others will undoubtedly occur after publi¬ cation. In citing applications of total quality in these companies, we have generally preserved their original names to clarify that the practices and results cited occurred under their original corporate identities.
SUPPORT MATERIALS FOR INSTRUCTORS The following support material is available from http://www.swlearning.com or the Thomson Learning Academic Resource Center at 800-423-0563. All instructor ancillaries are combined in the Instructor's Resource CD (ISBN: 0-324-20226-1).
Preface
• The Instructors' Manual—Prepared by author William Lindsay contains teaching suggestions and answers to all end-of-chapter questions, exercises, problems, and cases. • Power Point™ presentation slides—Prepared by author Jim Evans for use in lectures. New in this edition are Solved Problems in PowerPoint format to illus¬ trate quantitative techniques, prepared by author William Lindsay. Test Bank—Prepared by Darrell Radson of John Carroll University, the Test Bank includes true/false, multiple-choice, and short answer questions for each chapter. Exam View® computerized testing software allows instructors to create, edit, store, and print exams. ExamView also allows online test delivery. Possible Course Outlines Because the textbook material is comprehensive, it normally cannot be covered fully in one course. The textbook is designed to be flexible in meeting instructor needs. We have used it in both undergraduate courses and in managerially oriented MBA electives. We believe that undergraduate majors in industrial or operations management are best served by developing hands-on knowledge that they will be able to use in their entry-level jobs. Thus a typical course for these undergraduate students might emphasize the material in Parts 1 and 3, with some overview of the topics in Part 2. For MBAs, coverage of most of the first 9 chapters would be more appropriate for a quarter-long course, while much of Part 3 can be included in a semester-long course. At the University of Cincinnati, for example, we offer a 10-week MBA course that covers Chapters 1-9, and a shorter, 5-week course focused on the remaining chapters and Six Sigma. ACKNOWLEDGMENTS We are extremely grateful to all the quality professionals, professors, reviewers, and students who have provided valuable ideas and comments during the development of this and previous editions. For this sixth edition of The Management and Control of Quality, we received excellent feedback and suggestions from colleagues who partic¬ ipated in a focused survey. Our thanks and appreciation go to: Richard Benedetto, Merrimack College Victor L. Berardi, Kent State University, Stark Greg Blundell, Kent State University, Stark Gary Bragar, Bloomfield College Jon Burch, Trevecca Nazarene University Brenda J. Condrick, Western International University Deborah F. Cook, Virginia Polytechnic Institute and State University Robert A. Cornesky, Southern Wesleyan University Christine L. Corum, Purdue School of Technology Kazem Darbandi, California State Polytechnic University, Pomona David Doll, California State University, Hayward Craig G. Downing, Southeast Missouri State University Ellen J. Dumond, California State University, Fullerton Ahmad Elshennawy, University of Central Florida Robert F. Grant, Carthage College Marilyn M. Helms, Dalton State College
xxi
Preface
XXII
Anil Jambekar, Michigan Technological University William "Coty" Keller, St. Joseph's College, New York Henry W. Kraebber, Purdue University Frances Kubicek, Kalamazoo Valley Community College David Lewis, University of Massachusetts, Lowell Kevin Linderman, University of Minnesota Debra P. Maddox, Barry University Sara McComb, University Massachusetts, Amherst Jalane M. Meloun, Kent State University Muhammad Obeidat, Southern Polytechnic State University K. Praveen Parboteeah, University of Wisconsin, Whitewater Jim Pesek, Clarion University of Pennsylvania Darrell Radson, John Carroll University J. M. Thom, Purdue University John Todd, University of Arkansas Chiang Wang, California State University, Sacramento Don Wardell, University of Utah Geoff Willis, University of Central Oklahoma Many people deserve special thanks for their contributions to development and production of the book. Our regards go to senior acquisitions editor Charles McCormick, Jr., senior developmental editor Alice Denny, production editor Chris Sears, and designer Bethany Casey at Thomson Business and Professional Pub¬ lishing, and Richard Fenton, Mary Schiller, and Esther Craig, our previous editors at West Educational Publishing. Quality expert Joseph Juran was asked in an interview in 2002 what advice he would give to someone just starting out in quality today. Fie replied, "I would start out by saying 'Are you lucky!' Because I think the best is yet to be. In this current cen¬ tury, we are going to see a lot of growth in quality because the scope has expanded so much . . . away from manufacturing to all the other industries, including the giants: health care, education, and government." We will continue to do our best to improve this book in our quest for quality and to spread what we truly believe is a fundamentally important message to future gen¬ erations of business leaders. We encourage you to contact us at our e-mail addresses with any comments or improvement suggestions you may have. Please recall that instructor ancillaries may be requested from your Thomson publisher's representative, from http://evans.swlearning.com, or by calling the Aca¬ demic Resource Center at 800-423-0563 in the United States. James R. Evans ([emailprotected] ) William M. Lindsay ([emailprotected])
The Quality System Today, we generally do not hear about quality in business, except when things go wrong. Here is one example: "Spend $25,000 on a car that doesn't run the way you expect it to, and you get pretty angry. Spend $50,000 or $100,000, and you get really anSrY- Just listen to the anguished howls of Mercedes-Benz owners on Web sites as they vent about the latest mishap to afflict their Benzes. Depending on the model, the complaints range from faulty key fobs and leaky sunroofs to balky electronics that leave drivers and their passengers stranded. Regardless of the severity, a single sentiment runs through the gripes: this shouldn't be happening to a Mercedes." Sto¬ ries of successful organizations generally end up in publications dedicated to quality professionals, which basically "preach to the choir." We believe that less attention is paid to quality today as the result of two forces— a "good-news, bad-news" type of story. The good news is that the principles of quality that were new to many organizations in the early 1980s have become a common part of routine management practice; in other words, quality is so ingrained in the cultures of many organizations that managers and employees need not con¬ sciously think about it. As Mercedes' longtime CEO noted, "Quality is part of our heritage, one of our core values." The bad news is that without a conscious focus on it, it is easy for quality to slip by the wayside, as apparently happened at Mercedes. For many other organizations, quality is viewed as a short-term fix; when the hype and rhetoric passes, so do their quality efforts. Quality often still takes a backseat to economic pressures. Nevertheless, quality has not faded away, and will not fade away, simply because it works, with clear evidence that it improves the bottom line. Quality efforts are alive and well, perhaps under a different moniker in some organizations, and will remain an important part of a continual quest for improving performance across the globe. Joseph Juran, one of the most respected leaders of quality in the twentieth century, suggested that the past century will be defined by historians as the century of pro¬ ductivity. He also stated that the current century has to be the century of quality. "We've made dependence on the quality of our technology a part of life."1 As a member of the emerging generation of business leaders, you have an opportunity and a responsibility to improve the quality of your company and society, not just for products and services, but in everything you say and do. Part 1 introduces the basic concepts of quality. Chapter 1 discusses the history, definition, basic principles of quality, and the impact of quality on competitive
Part 1
2
Foundations of Quality Management
advantage and business results. Chapter 2 describes the role of total quality in dif¬ ferent types of organizations—manufacturing, service, health care, education, and government—and stresses the importance of taking a systems perspective of quality throughout an organization. Chapter 3 introduces the management philosophies on which modern concepts of quality are based, and managerial frameworks—the Mal¬ colm Baldrige Criteria for Performance Excellence, ISO 9000, and Six Sigma—that guide today's organizational approaches to quality improvement and performance excellence. These topics provide the foundation for the key quality principles and practices that are the subject of the remainder of the book.
ENDNOTES 1. Alex Taylor III, "Mercedes Hits a Pothole," Fortune, October 27, 2003,140-146.
2. Thomas A. Stewart, "A Conversation with Joseph Juran," Fortune, January 11,1999,168-169.
1
CHA|TE
R
Introduction to Quality The History and Importance of Quality The Age of Craftsmanship
QUALITY Profiles: Clarke American Checks, Inc., and Motorola, Inc. The Early Twentieth Century
Quality as a Management Framework Principles of Total Quality Infrastructure, Practices, and Tools
Quality and Competitive Advantage Quality and Business Results
Post-World War II
Three Levels of Quality
The U.S. "Quality Revolution"
Quality and Personal Values
Early Successes
QUALITY in Practice: The Evolution of Quality at Xerox: From Leadership Through Quality to Lean Six Sigma Quality in Practice: Bringing Total Quality Principles to Life at KARLEE
From Product Quality to Performance Excellence Disappointments and Criticism Current and Future Challenges
Defining Quality Judgmental Perspective
Review Questions
Product-Based Perspective User-Based Perspective
Discussion Questions Projects, Etc.
Value-Based Perspective
CASES
Manufacturing-Based Perspective Integrating Perspectives on Quality
Skilled Care Pharmacy A Tale of Two Restaurants A Total Quality Business Model
Customer-Driven Quality
Quality is by no means a new concept in modern business. In October 1887, William Cooper Procter, grandson of the founder of Procter & Gamble, told his employees, "The first job we have is to turn out quality merchandise that consumers will buy and keep on buying. If we produce it efficiently and economically, we will earn a profit, in which you will share." Procter's statement addresses three issues that are critical to managers of manufacturing and service organizations: productivity, cost, and quality. Productivity (the measure of efficiency defined as the amount of output achieved per unit of input), the cost of operations, and the quality of the goods and services that create customer satisfaction all contribute to profitability. Of these three determinants
3
Part 1
4
Foundations of Quality Management
of profitability, the most significant factor in determining the long-run success or failure of any organization is quality. High-quality goods and services can provide an organization with a competitive edge. High quality reduces costs due to returns, rework, and scrap. It increases productivity, profits, and other measures of success. Most importantly, high quality generates satisfied customers, who reward the orga¬ nization with continued patronage and favorable word-of-mouth advertising. To better understand the relationship among these factors, just consider Ford Motor Company. During the 1980s, Ford fought its way from the bottom of Detroit's Big Three automakers to the top of the pack through a concerted effort to improve quality and better meet customer needs and expectations. It quickly became a highly profitable business. However, on January 12, 2002, a newspaper headline read, "Ford to cut 35,000 jobs, close 5 plants." CEO William Ford is cited as stating "We strayed from what got us to the top of the mountain, and it cost us greatly. . . . We may have underestimated the growing strength of our competitors. There were some strategies that were poorly conceived, and we just didn't execute on the basics of our business." The article goes on to observe that Ford "has been dogged by quality problems that forced the recall of several new models, including the Explorer, one of the top money¬ makers."1 One of the key elements of Ford's 2002 Revitalization Plan was to "Continue Quality Improvements." In fact, the top two "vital few priorities" set by Ford's presi¬ dent for North America are "Improve quality" and "Improve quality"! If it were an easy task, there would be little need for this book. The mandate for focusing on quality is clear. In working with Chrysler Corporation (now Daimler-Chrysler) to improve quality, a vice president of the United Auto Workers (UAW) succinctly stated the importance of quality: Building—and maintaining—quality "No quality, no sales. No sales, no profit. No profit, into an organization's goods and ser¬ no jobs." vices, and more importantly, into the infrastructure of the organization itself, In this chapter we examine the notion of quality. is not an easy task. We discuss its history, its importance in business, and its role in building and sustaining competitive advantage. At the beginning of each chapter we profile two leading companies that have developed exemplary quality management practices (see the Quality Profiles on page 5). These examples will help you understand some of the key cultural issues that comprise the foundation of high-performing organizations. THE HISTORY AND IMPORTANCE OF QUALITY In a broad sense, quality assurance refers to any planned and systematic activity directed toward providing consumers with products (goods and services) of appro¬ priate quality, along with the confidence that products meet consumers' requirements. Quality assurance, usually associated with some form of measurement and inspection activity, has been an important aspect of production operations throughout history.2 Egyptian wall paintings circa 1450 b.c. show evidence of measurement and inspection. Stones for the pyramids were cut so precisely that even today it is impossible to put a knife blade between the blocks. The Egyptians' success was due to the consistent use of well-developed methods and procedures and precise measuring devices. The Age of Craftsmanship During the Middle Ages in Europe, the skilled craftsperson served both as manufac¬ turer and inspector. "Manufacturers" who dealt directly with the customer took con¬ siderable pride in workmanship. Craft guilds, consisting of masters, journeymen.
Chapter 1
Introduction to Quality
5
Quality Profiles
^roed/
_Q 0
C 0 0 0 0 > 4—*
E o
0) o c CO E Im
o
t: CD
CL
_0> O QC
“O
o o
o
0 ■H*
3
X 'Z
I
o 3 O
4-»
O X
O
x o
> 3 S o E ~ 0 a) ja J cn
o £ ^
4-*
e £ 0 0 — 0 .c 1 E ^ 0 -- 0 4-4
‘c/> —03 C T3 C § O 03
03
0 0 ^
x c 0
0 -a
0 0
X X
o
’0 4—' c
’0 0
p o
cn
CO -X
a-^ 0
> CD
11=
E
-5
o
C
0 CL 0 0
"0 0 X
03
C
0 X 0 CL X 0
■^00 03 _C =Z ^ 0 0 X
*o
00 0
c -s 0
Q. -2 E a)
X = J2
E
0 c
CO X
O 4—* 0 i_ 0 X
*D c 0
11 CT 0 03 X 0 0 X 0 1 3 4—< E CD
8 -
3 0 v E
e 0 CD 4-
E C O 0
E
° 3 JZ
c X 0 O-- — 0 0 JD _03 ■*“ 3 ©
CO
0 03 O
0 = 0 zr
.hz X
> >
0 -C 4-*
5
0 3 4-J o 0 4—
•- 2 ^ 0 £ 0 +3 O O £-§ C 0
t X
X* C O X 03
.E ° = “> 0 § 0
03 ,i= C 0
0 Q.
0.0
£.1 E x> 0
E
’0
o
^ 03
8. 2
0 0
0 —
■S o 2 — cd to aj 0
-C
03 C X jQ ^
.
S
o
©.
4-
it >
o
2
O
03 r_ 0 0
X
2
E O
c
8
c
2
0 0 0
E c E 03 03
E w E 0
+=
4—'
0 0 Q. 0 03 q r x .E o ^ t 0 4-“ 0 o 0X0
0 X L_ 0 O 0 O 0
4->
"4—1
0 ^ C
0 0 CL 03
E
E x
4-»
> -£ ^ 3 x 0
03 0 -E, S E 03 5^ 0 2 £ 0 co
0 0 03
03 O
E
c t
s 03
O00 4-* _C 0 0 4^ C 0X0 03 t > C 0 k® ■Fs
0 -Si!
811
8
0 X 0 4_> r_
V- C C O II 0 c CD 4-c C •2 1 0 cn > > 0 LX 0 .E > ^ x 0 0 J=- -X O O
° O ID 05 cn
2 -tn
5
C
O
c CO 0 ® t > l— 4—> 0 O > .9, O ‘FT
E 0 2 c ‘— 0 x n
= E .2
111
CO
o QC
Q. 0 4—• 0 0 03 03 3
3
0 0 X > % ” 0 X > C 03 o c x 0 X o 0 '00 c
E
8 8 8
8 0 ■a 1 E S 0 10 cn
1
0 0 0 0
s !> > o
cl 0 4-4 0 c 03 _0 t
o
2 03 C 0
'© 0 D O
Chapter 6
Human Resource Practices
CO CD 4-»
05 C c O 0 o
® 0
c
0 £ k. r
0 " -O
0 o
0 Q. O
0) "O o o
a 0
+-
3
-Q 'Z
JD (D ■g
c o o
3 O 0
b: g
E CD
TD CO 0)
0 0 O
0 73 JZ CO
>
E t O
o 05 c o
£3
o
c — CD C
w TS C
0 C 0 0 0 to t: O CL ^ J0 O O "O t o CO 4- CL C *4—
cl£
8 8 oc
4—1 0 k_ 3 O O 0
ii
_0 JD 0 a 0 b O a:
O 4~* 0 c 0
If
Q. O 0
0 3 Q. 0
0 4—* 0 o
S QCD Q.
0 0
2 0
•f 8
CD ~
E
’0
E Z5 O O “O
*4—* "0 0 0 0
E 0 > £ ■a L~ b co
CL f/5 o 03 0 05 O .b. c 0 C +3 0 0 Q_
JZ E 4—* ’4—< c ■O 0 0 ~G > JD
0 CL 0 0
c
“O
'05
0 -Q
H 0
4-» 0 0 JZ
O) c
.C
_0
JZ
d3 c
T3
m CO 0
0 "O 0
CD
£
^ JZ 0 4-» 0 c 0 .= -C o
1q CO
0 D
'0 S2 0 3 O 0
E |£ co £ 0 0 L t 0 .b
cc
0 0 = ’4—> 0
0 c o
0 X5 C — 00
_0
o o
®
'>
c — d) "O
o
■O
05 C
c 0 JI $
0 JZ
271
S 2
0 C 05 CD O £ if n
> Q L_ § CL r0 4-» | "O '0 0 05 4-» O _0 5 0 0 0 0 0 4—< C 3 O O 0 JD i_ 0 0 Q. 05
t > 0 c CL CD X 0 0 VVJ 0 0 CD
CD C i_ 0 c JZ — 0
tr 0 a CL 3 0
■g O T3 05 0 C Q- 0 0 JZ 0 £Z O 05 "0 4—' C Z3 0 > 4—1 0 0 3 CL 0 .E j= E O JZ 0 *4— O 1— JU 0 O O
k0 0 ts 4—* 0 c 0 E Cl 4—• O 'o .2 t JD 0 3 CL 0
0 ■£ 53 CL X 0
O
$
o ^cj CD
Q) A CO
£ CO
Z 0 O CC
k_ O 0
0 CL 0 0 0
O 0 LL
E P
0 _Q O (/)
0 k_ 0 JD mem
CSJ
Team
T3
D Q)CL
(0 CD
CC
3 ro
0O a3
272
Part 2
Quality in High-Performance Organizations
ronment with flexible practices, as shown in Table 6.3. Conventional HRM practices generally fall on the left side of each continuum. HRM choices that support a TQ envi¬ ronment are listed on the right-hand side pf the table. Day-to-day management activ¬ ities, which include how employees are selected and developed, how they are
Table 6.3 Choices in Designing Work Systems Planning Choices Informal Short term Explicit job analysis Job simplification Low employee involvement
. . . . .
Formal Long term Implicit job analysis Job enrichment High employee involvement
Staffing Choices Internal sources Narrow paths Single ladder Explicit criteria Limited socialization Closed procedures
. . . . . .
External sources Broad paths Multiple ladders Implicit criteria Extensive socialization Open procedures
Appraising Choices Behavioral criteria . Results criteria Purposes: Development, Remedial, Maintenance Low employee participation . High employee participation Short-term criteria . Long-term criteria Individual criteria . Group criteria Compensating Choices Low base salaries Internal equity Few perks Standard, fixed package Low participation No incentives Short-term incentives No employment security Hierarchical
High base salaries External equity Many perks Flexible package High participation Many incentives Long-term incentives High employment security High participation
Training and Development Choices Short term Narrow application Productivity emphasis Spontaneous, unplanned Individual orientation Low participation
Long term Broad application Quality-of-work-life emphasis Planned, systematic Group orientation High participation
Source: Adapted from R. S. Schuler, "Human Resource Management Practice Choices," in Readings in Personnel and Human Resource Management, 3d ed„ R. S. Schuler, S. A. Youngblood, and V. L. Huber, eds. (St. Paul, MN: West Publishing Company, 1988).
Chapter 6
Human Resource Practices
motivated at work, and how their performance is evaluated, can have a major impact on the success or failure of total quality efforts in an organization. In this section we address the most important elements in designing high-performance work systems that support a total quality focus. Work and Job Design Work design refers to how employees are organized in formal and informal units, such as departments and teams. Job design refers to responsibilities and tasks assigned to individuals. Both work and job design are vital to organizational effec¬ tiveness and personal job satisfaction. Unfortunately, managers often do not under¬ stand workers' needs. One research study found that the top five employee needs in the workplace are (1) interesting work, (2) recognition, (3) feeling "in" on things, (4) security, and (5) pay. Managers, however, believed pay to be number one. Many com¬ panies understand that the best way to influence job satisfaction and motivate workers is to make jobs more rewarding, which The design of work should provide can entail introducing variety into work (con¬ individuals with both the intrinsic sider the job rotation program at Sunny Fresh and extrinsic motivation to achieve Foods in the Quality Profiles), emphasizing the quality and operational performance importance and significance of the job, pro¬ objectives. viding more autonomy and empowerment, and giving meaningful feedback. An integrating theory that helps us understand how job design impacts motiva¬ tion, satisfaction, and organizational effectiveness was proposed by Hackman and Oldham.24 Their model, which has been validated in numerous organizational set¬ tings, is shown in Figure 6.4. The model contains four major segments: 1. 2. 3. 4.
Critical psychological states Core job characteristics Moderating variables Outcomes
Three critical psychological states drive the model. Experienced meaningfulness is the psychological need of workers to have the feeling that their work is a significant con¬ tribution to the organization and society. Experienced responsibility indicates the need of workers to be accountable for the quality and quantity of work produced. Knowl¬ edge of results implies that all workers feel the need to know how their work is evalu¬ ated and the results of their evaluation. Five core job characteristics have been identified as having an impact on the crit¬ ical psychological states: 1. Task significance: The degree to which the job gives the participants the feeling that they have a substantial impact on the organization or the world, for example, solving a customer's problem rather than simply filing papers 2. Task identity: The degree to which the worker can perceive the task as a whole, identifiable piece of work from start to finish, for example, building an entire component rather than performing a small repetitive task 3. Skill variety: The degree to which the job requires the worker to use a variety of skills and talents, for example, physical skills in machining a part and mental skills in using a computer to track quality measurements 4. Autonomy: The degree to which the task permits freedom, independence, and persona] control to be exercised over the work, for example, being able to stop a production line to solve a problem
273
Part 2
274
Quality in High-Performance Organizations
Figure 6.4 Hackman and Oldham Work Design Model
Skill variety Experienced meaningfulness of
Task identity
the work Task significance
High internal work motivation High “growth” satisfaction
Experienced Autonomy-►
responsibility for
High general
outcomes of the work
job satisfaction
Knowledge of the Feedback from job-► actual results of the work activities
1
High work effectiveness
Moderators: 1. Knowledge and skill 2. Growth need strength 3. “Context” satisfactions
Source: J. Richard Hackman and Greg R. Oldham, Work Redesign (figure 4.6 from p. 90). © 1980 by AddisonWesley Publishing Co., Inc. Reprinted by permission of Addison Wesley Longman.
5. Feedback from the job: the degree to which clear, timely information about the effectiveness of performance of the individual is available, not only from super¬ visors, but also from measurements that the worker might take directly. Quality is related in a primary or secondary sense to all five of these core job char¬ acteristics. Quality of a product or service is undoubtedly increased by a worker's dedicated application of skills, which is enhanced by task identity and a feeling of task significance. More directly, quality of work is enhanced by a job design that incorporates autonomy and feedback relating to quality characteristics. The key out¬ comes of high general job satisfaction and high work effectiveness can then be seen as results that define and reinforce excellent quality. As an example illustrating characteristics of the Hackman and Oldham model, consider the case of workers in a small Delaware firm that produces space suits for astronauts. The work requires a great deal of handcrafting, using conventional sewing machinery as well as high technology in testing the suits for proper func¬ tioning. Task significance and task identity are evident in the workers' ability to see the job's extreme importance and its fit into the complete unit (a space suit for an individual astronaut). Skill variety and autonomy are somewhat limited because con¬ ventional sewing techniques must be used and rigid specifications must be precisely followed. However, other motivating aspects of the job may compensate for the lack of these characteristics. Feedback on results is timely and individualized. Compre¬ hensive testing and inspection of the space suits is performed to assure that no defec¬ tive units are produced.
Chapter 6
Human Resource Practices
Several common approaches to work design—job enlargement, job rotation, and job enrichment—are supported by this model. IBM was apparently the first user of job enlargement, in which workers' jobs were expanded to include several tasks rather than one single, low-level task. This approach reduced fragmentation of jobs and generally resulted in lower production costs, greater worker satisfaction, and higher quality, but it required higher wage rates and the purchase of more inspection equipment. Job rotation is a technique by which individual workers learn several tasks by rotating from one to another. The purpose of job rotation is to renew interest or motivation of the individual and to increase his or her complement of skills. How¬ ever, several studies showed that the mam benefit was to increase workers' skills but that little, if any, motivational benefit could be expected.25 Finally, job enrichment entails "vertical job loading" in which workers are given more authority, responsi¬ bility, and autonomy rather than simply more or different work to do. Garvin pre¬ sents an interesting example of how Japanese managers in the air-conditioning industry view job enrichment as important to quality.26 In Japan, newly hired workers are trained so that they can do every job on the line before eventually being assigned to only one job. Training frequently requires 6-12 months, in contrast to the standard training time of one to two days for newly hired production workers in U.S. airconditioning companies. The advantage of this "enriched" training is that workers are better able to track a defect to its source and can frequently suggest remedies to problems because they understand the entire process from start to finish. Job enrich¬ ment has been used successfully in a number of firms, notably AT&T, which experi¬ enced better employee attitudes and performance, as well as Texas Instruments, IBM, and General Foods. Employee Involvement Tom Peters suggested involving everyone in everything, in such activities as quality and productivity improvement, measuring and monitoring results, budget develop¬ ment, new technology assessment, recruiting and hiring, making customer calls, and participating in customer visits.27 Employee involvement (El) refers to any activity by which employees participate in work-related decisions and improvement activi¬ ties, with the objectives of tapping the creative energies of all employees and improving their motivation. Pete Coors, CEO of Coors Brewing, explained it simply, "We're moving from an environment where the supervisor says, 'This is the way it is going to be done and if you don't like it, go someplace else,' to an environment where the supervisor can grow with the changes, get his troops together and say, 'Look, you guys are operating the equipment, what do you think we ought to do?'"28 The continuum of El approaches is sum¬ marized in Table 6.4. As total quality matures El approaches can range from simple in an organization, higher levels of employee sharing of information or providing involvement are evident. One of the most input on work-related issues and making suggestions to self-directed prominent employee involvement processes responsibilities such as setting goals, has been GE's "Work-Out" program.29 Em¬ making business decisions, and solv¬ ployees are encouraged to get together in a ing problems, often in crossseries of meetings to discuss reports, meetings, functional teams. measurements, and approvals in their work area or department. The meetings are facili¬ tated by an outside leader, but supervisors are forbidden to attend, except for a brief opening appearance, until the last day of a three-day session. At the final Work-Out session, the supervisor, and often, his or her boss, are at the front of the room, with no
276
Part 2
Quality in High-Performance Organizations
Table 6.4 Levels of Employee Involvement
Level
Action
Primary Outcome
1.
Information sharing
Managers decide, then inform employees
Conformance
2.
Dialogue
Managers get employee input, then decide
Acceptance
3.
Special problem solving
Managers assign a one-time problem to selected employees
Contribution
4.
Intragroup problem solving
Intact groups meet weekly to solve local problems
Commitment
5.
Intergroup problem solving
Cross-functional groups meet to solve mutual problems
Cooperation
6.
Focused problem solving
Intact groups deepen daily involvement in a specific issue
Concentration
7.
Limited self-direction
Teams at selected sites function full time with minimum supervision
Accountability
8.
Total self-direction
Executives facilitate self-management in an all-team company
Ownership
Source: Copyright © Jack D. Orsburn, Linda Moran, Ed Musselwhite, and John H. Zenger, Self-Directed Work Teams! Burr Ridge, IL: Business One Irwin, 1990), 34. Reproduced with permission of The McGraw-Hill Companies.
idea of what has been discussed during the previous two days. The supervisor can only respond to items that the employees recommend in one of three ways: 1. Agree on the spot to implement the proposal. 2. Say no to the proposal. 3. Ask for more information. Typically, more than 80 percent of the Work-Out recommendations received an immediate answer. For example, Armand Lauzon, head of plant services at GE Air¬ craft Engines factory in Lynn, Massachusetts, was confronted with 108 proposals at the end of a Work-Out session by his employees. He said yes to 100 recommendations on the spot, including one in which an employee had sketched a design for protective shields for machines on a brown paper bag. The employee asked whether his group could bid on the work. They got the bid when they quoted a cost of $16,000 versus an outside vendor's proposed cost of $96,000! El initiatives are by no means new.31' Many programs and experiments have been implemented over more than 100 years by industrial engineers, statisticians, and behavioral scientists. Early attempts influenced modern practices considerably. Unfortunately, these approaches lacked the complementary elements of TQ, such as a customer orientation, top management leadership and support, and a common set of tools for problem solving and continuous improvement. El is rooted in the psychology of human needs and supported by the motiva¬ tion models of Maslow, Herzberg, and McGregor. Employees are motivated through exciting work, responsibility, and recognition. El provides a powerful means of achieving the highest order individual needs of self-realization and fulfillment. Thus,
Chapter 6
Human Resource Practices
employee involvement should begin with a personal commitment to quality, as we discussed in Chapter 1. El offers many advantages over traditional management practices: • Replaces the adversarial mentality with trust and cooperation • Develops the skills and leadership capability of individuals, creating a sense of mission and fostering trust • Increases employee morale and commitment to the organization • Fosters creativity and innovation, the source of competitive advantage • Helps people understand quality principles and instills these principles into the corporate culture • Allows employees to solve problems at the source immediately • Improves quality and productivity31 One of the easiest ways to involve employees on an individual basis is the sug¬ gestion system. An employee suggestion system is a management tool for the sub¬
mission, evaluation, and implementation of an employee's idea to save cost, increase quality, or improve other elements of work such as safety. Companies typically reward employees for implemented suggestions. At Toyota, for instance, employees generate nearly 3 million ideas each year—an average of 60 per employee—of which 85 percent are implemented by management. Suggestion systems are often tied to incentives. Wainwright Industries developed a unique and effective approach that has been benchmarked extensively.32 Suggestion programs were viewed as neither systematic nor continuous, and not woven into the fabric of daily operations. Their approach was designed to overcome these shortcomings in the following ways: • Focusing employees on small, incremental improvements within their own areas of responsibility and control • Recognizing all employees for their level of participation regardless of the value of the improvement • Scaling team-based improvement efforts in a way that minimizes downtime and provides people with the tools and techniques to produce successful out¬ comes • Positioning supervisors as the catalyst for cultural change through a coaching and support role in the employee involvement and improvement process The process contains two main components: individual implemented improve¬ ments and team-based system improvements. Rather than submitting suggestions for someone else to approve and implement, employees are provided with training and given the responsibility to take the initiative to make improvements on their own without prior approval within the scope of their main job responsibilities. Upon making improvements, they complete a form to document what they have done and present it to the supervisors, whose role is not to approve or disapprove, but to acknowledge the improvement and to point out any issues that the employee needs to understand. All forms submitted during the week are placed into a random drawing for some type of award determined by the individual unit. At the end of each quarter, every individual who met his or her goal of implemented improve¬ ments receives some type of valued recognition. The team-based approach breaks large initiatives into smaller manageable projects. Breaking down large tasks allows employees to understand how their individual jobs fit into the big picture and maxi¬ mizes participation reduces time requirements for any particular employee. Wain¬ wright was able to cite more than 50 implemented improvements per employee per year, far exceeding those of most American and Japanese companies.
277
278
Part 2
Quality in High-Performance Organizations
Fostering employee creativity has many benefits. Thinking about solutions to problems at work makes even routine work enjoyable; writing down the suggestions improves workers' reasoning ability and writing skills. Satisfaction is the by-product of an implemented idea and a job made easier, safer, or better. Recognition for sug¬ gestions leads to higher levels of motivation, peer recognition, and possible monetary rewards. Workers gain an increased understanding of their work, which may lead to promotions and better interpersonal relationships in the workplace. Table 6.5 sum¬ marizes strategies that can foster the success of suggestion systems. Empowerment
Empowerment requires, as the management philosophy of Wainwright Industries states, a sincere belief and trust in people. A survey by Annandale, Virginia-based MasteryWorks Inc. concluded that employees leave their organizations because of trust, observing that "Lack of trust was an issue with almost every person who had left an orga¬ Empoiverment simply means giv¬ nization."33 ing people authority—to make deci¬ Examples of empowerment abound. Workers sions based on what they feel is in the Coors Brewery container operation give right, have control over their work, each other performance evaluations, and even take risks and learn from mistakes, screen, interview, and hire new people for the and promote change. line. At Motorola, sales representatives have the
Table 6.5 Success Factors for Suggestion Systems 1. Ensure that management, first and foremost, is involved in the program. Involvement should begin at the top and filter down through all levels until all employees participate. 2. Push decision making regarding suggestion evaluation to lower levels. 3. Gain union support by pledging no layoffs due to productivity gains from adopted suggestions. 4. Train everyone in all facets of the suggestion system. Improve problem-solving capability by promoting creative problem solving through the use of the seven basic statistical tools. 5. Resolve all suggestions within one month. 6. Encourage all suggestors to personally describe their idea to a supervisor, engi¬ neer, or manager. 7.
Promote pride in work, and quality and productivity gains from suggestions, rather than the big cash awards if possible. 8. Remove ceilings on intangible suggestion awards. Revise evaluations of intan¬ gible suggestions to value them more on par with tangible suggestions. 9. Eliminate restrictions prohibiting suggestions regarding a worker's immediate work area. 10. Continuously promote the suggestion program, especially through supervisor support. 11. Trust employees enough to make allowances for generation, discussion, and submittal of suggestions during work hours. 12. Keep the program simple.
Source: Muse and Finster, "A Comparison of Employee Suggestion Systems in Japan and the USA," University of Wisconsin Working Paper (1989).
Chapter 6
Human Resource Practices
authority to replace defective products up to six years after purchase, a decision that used to require top management approval. A Corning Glass plant replaced 21 dif¬ ferent jobs with one "specialist" job and gave employee teams broad authority over production scheduling and division of labor. The need to empower the entire workforce in order for quality to succeed has long been recognized. Juran wrote that "ideally, quality control should be delegated to the workforce to the maximum extent possible."34 Five of Deming's 14 Points relate directly to the notion of empowerment: Point Point Point Point Point
6: Institute training. 7: Teach and institute leadership. 8: Drive out fear. Create trust. Create a climate for innovation. 10: Eliminate exhortations for the workforce. 13: Encourage education and self-improvement for everyone.35
These points suggest involving employees more directly in decision-making processes, giving them the security and confidence to make decisions, and providing them with the necessary tools and training. Empowered employees must have the wisdom to know what to do and when to do it, the motivation to do it, and the right tools to accomplish the task.36 These require¬ ments may mean significant changes in work systems, specifically, the following: • Employees be provided education, resources, and encouragement. • Policies and procedures be examined for needless restrictions on the ability of employees to serve customers. • An atmosphere of trust be fostered rather than resentment and punishment for failure. • Information be shared freely rather than closely guarded as a source of control and power. • Workers feel their efforts are desired and needed for the success of the organi¬ zation. • Managers be given the required support and training to adopt a" hands-off" leadership style. • Employees be trained in the amount of latitude they are allowed to take. For¬ mulating decision rules and providing role-playing scenarios are excellent ways of teaching employees.37 Empowerment also means that leaders and managers must relinquish some of the power that they previously held. This power shift often creates management fears that workers will abuse this privilege. Flowever, experience shows that front-line workers generally are more conservative than managers. For example, companies that have empowered employee groups to evaluate performance and grant pay raises to their peers have found that they are much tougher than managers were. Empowerment gives managers new responsibilities. They must hire and develop people capable of handling empowerment, encourage risk taking, and recognize achievements. Giving employees information about company finances and the finan¬ cial implications of empowered decisions is also important. At DuPont's Delaware River plant, management shares cost figures with all workers.38 By sharing this information, management believes that workers will think more for themselves and identify with company goals. To help employees make decisions on issues affecting production, a department manager at the Eastman Chemical plant in Texas supplied operators with a daily financial report that showed how their decisions affected the bottom line. As a result, department profits doubled in four months
279
280
Part 2
Quality in High-Performance Organizations
and quality improved by 50 percent as employees began suggesting cost-saving improvements.39 Empowerment can be viewed as vertical teamwork between management and labor. It builds confidence in workers by showing them that the company has confi¬ dence in their ability to make decisions on their own. It generates commitment and pride. It also gives employees better experience and an opportunity to advance their careers. It benefits customers who buy the organization's products and services. For instance, empowered employees can often reduce bureaucratic red tape that cus¬ tomers encounter—such as seeking a supervisor's signature—which makes cus¬ tomer transactions speedier and more pleasant. John Akers, former chairman of IBM, said, "Empowering our employees and inculcating a sense that everyone owns his or her piece of the business not only unleashes the talent and energy of our people, but also flattens the organization and reduces stifling bureaucracy."40 Even though many workers prefer an empowered workplace to the old style of narrowly defined tasks, empowerment is not for everyone.41 One worker at Eaton Corporation hated the idea of being her own boss and its associated responsibilities such as fixing broken machines and having to learn a wide variety of jobs, and left after nine months. This example suggests that selecting the right people for a partic¬ ular work environment is an important task. Recruitment and Career Development
Motorola ties recruitment and selection activities to results in order to gauge the quality of its recruiting efforts as it strives for TQ at every level.42 The recruiting department is measured by a new quality-oriented criterion: success of recruits on the job. Instead of using the old measure of how much it costs to hire each recruit, recruiters are now mea¬ sured on whether new hires were well trained coming into the company, brought in at the right salary level, or left the company after the first six months for a better job. Based on these and other data, the department decided it had to increase, rather than decrease, the amount spent on each recruit. Thus, in recruiting activities. Motorola plans and sets objectives for recruiting, charts progress over time in order to reduce "defects" in the hiring process, and determines whether the "output" of the process (excellent employees) is under control, rather than simply measuring inputs (dollars per recruit). Other major companies like Procter & Gamble seek entry-level college graduates who understand total quality principles. They specifically want their new employees to think in terms of creating quality and value for consumers, to understand their cus¬ Meeting and exceeding customer tomers and needs, and to work toward results expectations begins with hiring the despite obstacles. right people whose skills and atti¬ tudes will support and enhance the Customer-contact employees make up one of organization's objectives. the fastest-growing segments of the workforce. Limited availability of people with the skills to perform complex, rapidly changing jobs is forcing HRM managers to rethink their selection strategies. Traditional hiring practices have been based on cognitive or tech¬ nical rather than interpersonal skills. The criterion is now shifting to attributes such as enthusiasm, resourcefulness, creativity, and the flexibility to learn new skills rapidly. The internal customer concept suggests that every employee needs good interper¬ sonal skills. Even technical skill requirements are changing; to apply quality principles on the job, all workers must have basic mathematics and logical-thinking abilities. To ensure that job candidates have the requisite skills, new approaches, such as psycho¬ logical testing and situational role playing, are now being used in the hiring process.
Chapter 6
Human Resource Practices
Career development is also changing because of TQ. As managerial roles shift from directing and controlling to coaching and facilitating, managers, who must deal with cross-functional problems, benefit more from horizontal movement than from upward movement in narrow functional areas. Flatter organizations limit promotion opportunities. Thus, career development expands learning opportunities and creates more challenging assignments rather than increasing spans of managerial control. Training and Education
Companies committed to TQ invest heavily in training and education, recognizing that such investments add value to organizational capabilities. The leaders in quality—Deming, Juran, and Crosby—actively promoted quality training and edu¬ cation. Two of Deming's 14 Points, for example, are devoted to these issues. Xerox Business Products and Systems, for instance, invested more than $125 million in quality training. Customer service representatives at FedEx receive five weeks of training before they ever speak unsupervised with a customer. Even an 18-employee digital printing company in Reykjavik, Iceland—Umslag, ehf—spends 4 percent of total wages on training, which includes training in equipment repair to reduce the need to bring in expertise from Holland, languages to support growth in interna¬ tional business, and personal interests, such as computer programming, that can ben¬ efit the business.43 Training is one of the largest initial costs in a total quality initiative. Not surpris¬ ingly, it is one in which many companies are reluctant to invest. However, research indicates that companies that spend heavily on training their workers outperform com¬ panies that spend considerably less, as measured on the basis of overall stock market returns. Even if companies make the investment, they often take great pains to measure the benefits against the costs. Motorola used to calculate returns, but no longer. Its man¬ agement knows that the benefits of quality-based training outweigh the costs by at least 30 to 1. Training and education have become an essential responsibility of HRM depart¬ ments in TQ organizations, particularly as empowered employees require new knowl¬ edge and skills, which should not have to be cost-justified. Training generally includes quality awareness, leadership, project management, communications, teamwork, problem solving, interpreting and using data, meeting customer requirements, process analysis, process simplification, waste reduction, cycle time reduction, error-proofing, and other issues that affect employee effectiveness, effi¬ ciency, and safety. Education needs might also include basic skills, such as reading writing, language, mathematics, or computer skills. Employees at Xerox learn a range of techniques, from the basic quality improvement tools through benchmarking. Motorola employees learn statistical methods and defect reduction approaches. Solec¬ tron Corporation, with a large multicultural workforce in its U.S. facilities, offers Eng¬ lish as a second language courses, and training in communications, interpersonal skills, and technical manufacturing skills, all with bilingual trainers. In a total quality environment, employees need to understand the goal of cus¬ tomer satisfaction, to be given the training and responsibilities to achieve this goal, and to feel that they do indeed make a difference. For example, at the Coors Brewing Company in Golden, Colorado, the customer satisfaction improvement program is focused on giving employees the appropriate skills, and on creating the environment in which employees have one responsibility and one hoped-for result: to satisfy and delight their customers, especially internal customers. Coors engaged in a massive training program to learn TQ principles, and then restructured its organization sys¬ tems (compensation, evaluation, and so on) to support the new effort. The company
281
282
Part 2
Quality in High-Performance Organizations
succeeded in developing in its employees a passion for their jobs and pride in their work, which translated into measurable improvements in productivity, a remarkably low turnover rate, and the delivery of quality product and service throughout the system.44 Some employees, such as customer-contact Customer needs and strategic direcpersonnel typically need a higher level of tions should drive training stratetraining in behavioral topics than manufacturing gies. engineers, who may need advanced statistical skills. FedEx, with its focus on efficiency and customer service, trains workers in team development and people issues. At IBM Rochester, managers tell the education department what they need, and programs are designed to meet those needs. By treating the training function as an internal supplier, the time taken to deliver training programs has been reduced from five days to two. Many large companies have formal training departments, whose systems and approaches evolved along with their overall quality systems. Education and training can be delivered in a variety of ways, including on-the-job or traditional classroom environments. Today, computer-based and distance learning education are becoming increasingly popular. Training can also be accomplished through developmental assignments within or outside the organization. Specific approaches vary by company. In some, managers train their workers directly in a top-down fashion; this approach was pioneered by Xerox, beginning with the CEO, David Kearns himself, during their transition to total quality. Others use self-paced methods employing advanced technology. The FedEx Quality Academy, established in 1991, uses a televi¬ sion network that broadcasts courses in a just-in-time fashion at the employees' work site. It also has a network of interactive video instruction, consisting of 1,200 work¬ stations at 700 locations. More than 2,000 course titles are available for self-paced instruction. The Quality Academy tracks test scores, pass rates, and time spent online.45 Honda of America uses interactive computer-based training modules on dedicated workstations in the plant.46 Smaller companies often use outside consul¬ tants. The content should be customized to the company's needs; "packaged" semi¬ nars are often a waste of time. Continual reinforcement of lessons learned in training programs is essential. Many companies send employees to courses, but then allow the knowledge to slip away. New knowledge can be reinforced in several ways. Motorola uses on-the-job coaching to reinforce training; The Ritz-Carlton has follow-up sessions to monitor instructional effectiveness. The Ritz-Carlton holds a "quality lineup" briefing session each day in every work area. During these sessions, employees receive instructions on achieving quality certification within the company. Work area teams set the quality certification performance standards of each position. Finally, companies need am approach for eval¬ uating training effectiveness. The Ritz-Carlton requires employees to pass written and skill demonstration tests. Other companies use on-the-job evaluation or tests in simu¬ lated work environments. Many measure behavior and attitude changes. However, the true test of training effectiveness is results. By establishing a linkage between training and results (see the discussion of interlinking in Chapter 8), companies can show the impact on customer satisfaction and also identify gaps in training. Compensation and Recognition
Without willing, sustained, individual effort and coordinated teamwork focused on meeting organizational goals, TQ is an impossible dream. However, when organiza¬ tions ask employees to assume new challenges and responsibilities, the question
Chapter 6
Human Resource Practices
Compensation and recognition refer to all aspects of pay and reward, including promotions, bonuses, and recognition, either monetary and nonmonetary or individual and group.
283
"What's in it for me?" ultimately gets asked. Extrinsic and intrinsic rewards are the key to sustained individual efforts.
Compensation is always a sticky issue, closely tied to the subject of moti¬ vation and employee satisfaction. Although money can be a motivator, it often causes employees to believe they are being treated unfairly, and forces managers to deliver negative messages. Eventually, it diminishes intrinsic motivation and creates win-lose situations. The objectives of a good compensation system should be to attract, retain, and not demotivate employees. Other objectives include reducing unexplainable vari¬ ation in pay (think about Deming's principles) and encouraging internal cooperation rather than competition. Most companies still use traditional financial measures, such as revenue growth, profitability, and cost management, as a basis for compensation; more progressive organizations use quality measures such as customer satisfaction, defect prevention, and cycle time reduction to make compensation decisions. Many TQ-focused companies now base compensation on the market rate for an individual with proven capabilities, and then make adjustments as capabilities are increased, along with enhanced responsibilities, seniority, and business results. For example. General Motors' Powertrain Division, influenced strongly by Denting, decoupled compensation from performance appraisals. Compensation is determined from a "maturity curve" that considers an individual's seniority, level of expertise, and market for his or her services. Peers and subordinates have input as to an indi¬ vidual's rating on this curve. Distinctions based on contributions are limited to truly exceptional individuals. This example is an exception; few companies have elimi¬ nated merit ratings from their salary systems. Many companies link compensation to company track records, unit performance, team success, or individual achievement.47 At Kaiser Aluminum, such performancebased compensation incentives led to an 80 percent improvement in productivity and 70 percent decrease in poor quality costs over five years.48 Team-based pay and gain¬ sharing, an approach in which all employees share savings equally, are gaining in popularity and importance. Compensation for individuals is sometimes tied to the acquisition of new skills, often within the context of a continuous improvement pro¬ gram in which all employees are given opportunities to broaden their work-related competencies. However, legal restrictions in federal wage and hour laws make it dif¬ ficult to implement some of these approaches. At a 1999 hearing of a subcommittee of the House Committee on Education and the Workforce, Pam Farr, a management consultant with the Cabot Advisory Group who had previously worked for Marriott Corporation as a human resources executive, testified: Compensation
A recent survey by William M. Mercer indicated that just 24 percent of large and midsize companies use team-based incentive pay. For companies that have chosen to implement team incentive pay programs, however, the results are overwhelmingly positive. A recent study by the Hay Group indi¬ cated that team-based and gainsharing plans are the most effective pro¬ grams to help improve employee performance and satisfaction. A General Accounting Office study indicated that such programs significantly improve employer-employee relations, and reduce grievances, absenteeism and turnover. By removing the impediments to team-based pay systems, Con¬ gress will facilitate employee pay increases, employee work satisfaction and encourage productivity increases,49
284
Part 2
Quality in High-Performance Organizations
Nucor Corporation, one of the nation's largest steel producers, is well-known for having succeeded in attacking quality, productivity, participation, and compensation issues.50 Nucor has more than 6,000 employees in plants in the United States and had annualized sales in excess of $4 billion in mid-1997. All employees, from the presi¬ dent on down, have the same benefits; the only differences in individual pay are related to responsibilities. Workers at Nucor's five nonunion steel mills earn base hourly rates that are less than half of the going rate for unionized steelworkers. Nucor uses pay incentives designed around groups of 40 to 50 workers, including secretaries and senior managers. They offer four basic compensation plans: 1. Production Incentive Plan. Employees involved directly in manufacturing are paid weekly bonuses on the basis of production of their work groups, which range from 20 to 40 workers each. These productivity and quality bonuses are based on the number of tons of steel of acceptable quality produced by a given production team. The formulas are nondiscretionary, based upon established production goals, and can average 80-150 percent of the base wage. This plan creates pressure for each individual to perform well, and in some facilities, is tied to attendance and tardiness standards. No bonus is paid if equipment is not operating, thus creating a strong emphasis on maintaining equipment in top operational condition at all times. The bonuses are paid every week to reinforce motivation. The average worker at Nucor earns several thousand dollars per year more than the average worker in the industry, while the company is able to sell its steel at competitive worldwide market prices. 2. Department Manager Incentive Plan. Department managers earn incentive bonuses paid annually based primarily on the return on assets of their facility. 3. Nonproduction and Nondepartment Manager Incentive Plan. Participants include accountants, engineers, secretaries, and other employees. The bonus is based on the facility's return on assets. Each month every operation receives a report showing progress, which is posted in the employee cafeteria or break area to keep employees appraised of their expected bonus levels throughout the year. 4. Senior Officers Incentive Plan. Senior officers do not receive profit sharing, pension, discretionary bonuses, or retirement plans. A significant part of their compensation is based upon Nucor's return on stockholder's equity above a certain minimum earnings. If Nucor does well, compensation is well above average, as much as several times base salary. If the company does poorly, com¬ pensation is limited to base salary, which is below the average pay at compa¬ rable companies. The company was producing a ton of steel for less than half the average costs of a U.S. steel company.51 For example, in September 1997, when other steel companies were attempting to raise prices for steel, Nucor announced that it was cutting the price of cold-rolled steel, one of the most widely used product lines, by 7 percent.52 Nucor required fewer than four hours of labor per ton, Japanese companies required about five hours per ton, and other U.S. mills averaged more than six hours per ton. This comparison illustrates the use and benefits of team-based pay policies. During downturns, managers at Nucor frequently find that their bonuses are cut, even while hourly workers continue to receive theirs, based on production rates. One difficult year, Nucor cut salaries for its 12 top executives by 5 percent and froze wages for its 3,500 employees. However, despite the tough times, it maintained their policy of no layoffs as it had throughout the history of the current company. The next year, when the United Steelworkers Union signed a contract to reduce wages and benefits in order to improve the competitiveness of the basic steel industry, Nucor announced
Chapter 6
Human Resource Practices
285
a 5 percent wage increase. More about the Nucor story can be found on its Web site at http://www.nucor.com. Special Recognition and Rewards Special recognition and rewards can be monetary or nonmonetary, formal or informal, individual or group. These rewards might include trips, promotional gifts, clothing, time off, or special company-sponsored awards and events. Most importantly, rewards should lead to behaviors that increase customer satisfaction. A Conference Board study found that a combination of cash and noncash recognition works better for clerical and hourly workers than for man¬ agers and professional/technical employees; for these groups, compensation-based incentives such as stock options are more successful.53 As an example, in October 1994, Continental Airlines, new CEO Gordon Bethune calculated that late and can¬ celed flights were costing the company $6 million per month to put passengers on rival airlines or send them to hotels. He declared that if Continental ranked among the top three airlines for on-time performance in any month, he would split half the savings (about $65 per person) with all nonexecutive employees. Within two months. Continental was first. To ensure that the bonuses made a vivid impression, Bethune issued the checks separately and traveled around the country to distribute thousands of them personally. The behavioral changes are best illustrated by a story executives like to retell. A catering truck pulls up to a plane but is 10 meals short. In the old days, the flight attendant would have told the driver to get the extra meals while the plane sat at the gate for 40 minutes. The newly gung ho flight attendant, however, crisply tells the Recogitition provides a visible means of promoting quality efforts and catering guy not to screw up again and shuts telling employees that the organiza¬ the cabin door. The plane pushes back on tion values their efforts, which stim¬ schedule, and she finds a bunch of investment ulates their motivation to improve. bankers and offers them free liquor in place of the meal.54 Employees should contribute to the company's performance and recognition approaches. L.L. Bean, for example, gives dinners or certificates exchangeable for mer¬ chandise. Winners of "Bean's Best Awards" are selected by cross-functional teams based on innovative ideas, exceptional customer service, role modeling, expertise at their jobs, and exceptional management ability.55 Certain key practices lead to effective employee recognition and rewards: • Giving both individual and team awards. At The Ritz-Carlton, individual awards include verbal and written praise and the most desirable job assignments. Team awards include bonus pools and sharing in the gratuity system. Many compa¬ nies have formal corporate recognition programs, such as IBM's Market Driven Quality Award for outstanding individual and team achievements in quality improvement, or the Xerox President's Award and Team Excellence Award. • Involving everyone. Recognition programs involve both front-line employees and senior management. Westinghouse had a Wall of Fame to recognize quality achievers at each site. Solectron rewards groups by buying lunch for entire divi¬ sions and bringing in ice cream for everyone in the plant. A Monsanto Com¬ pany chemical plant ties worker bonuses to results at individual units and rewards workers for helping to prevent accidents.56 What is particularly inter¬ esting is that different programs exist in different Monsanto plants—all devel¬ oped with the participation of workers. Bonus plans that failed had been ones decreed by corporate headquarters, rather than those formulated in coopera¬ tion with employees.
W
Airlines
286
Part 2
Quality in High-Performance Organizations
• Tying rewards to quality based on measurable objectives. Leading companies recog¬ nize and reward behavior, not just results. Zytec rewarded employees for par¬ ticipating in the suggestion program by providing cash awards for each implemented suggestion. A group of peers selected the best improvement ideas each month, which were also rewarded with cash. Many rewards were linked to customer satisfaction measures. Awards that conflicted with quality values were modified or eliminated. Continuous feedback reinforced good perfor¬ mance and identified areas for improvement. When Custom Research, Inc. attains a specific corporate goal, the entire company is taken on a trip to desti¬ nations such as San Francisco and Disney World! • Allowing peers and customers to nominate and recognize superior performance. Texas Instruments, for example, has a Site Quality Award to recognize the top 2 per¬ cent on the basis of peer nomination. Employees at FedEx who receive favorable comments from a customer are automatically nominated for the Golden Falcon Award. Recipients chosen by a review committee receive a gold pin, a congratu¬ latory call from the CEO, recognition in the company newsletter, and 10 shares of company stock. AT&T Universal Card Services' World of Thanks award con¬ sisted of a globe-shaped pad of colored paper with "Thank You" written in dif¬ ferent languages. Anyone in the company could write a message of thanks to someone else. In four years, employees used more than 130,000 of them!57 • Publicizing extensively. Many companies recognize employees through newslet¬ ters, certificates and pins, special breakfasts or luncheons, and annual events such as competitions. Motorola, for example, developed a worldwide Total Cus¬ tomer Satisfaction (TCS) team competition. Approximately half of Motorola's 142,000 employees are on teams that compete locally, regionally, and interna¬ tionally to attend the final one-day, corporate-wide competition held at a resort each year. The 1996 competition included 24 teams from eight countries. Teams were scored on such criteria as project selection, teamwork, analysis techniques, remedies, results, institutionalization (permanence, deployment, and team growth from the project), and presentation. Corporatewide results over eight years have been impressive, with an estimated savings of $2.4 billion per year.58 • Making recognition fun. Domino's Pizza stages a national Olympics, in which teams from the company's three regions compete in 15 events based on 15 job cat¬ egories, such as doughmaking, driving, answering the telephone, and delivery. Winners, standing on platforms while the Olympic theme is played, receive medals, checks, and other forms of recognition. The finals are broadcast live to commissaries around the country. Domino's Olympics provides an excellent way to benchmark efforts throughout the corporation; winners attend three days of discussion with upper management to talk about what's good about the com¬ pany, what needs improvement, and how those improvements can be made.59 Health, Safety, and Employee Well-Being
Because employees are key stakeholders of any organization, their health, safety, and overall well-being are important factors in the work environment. Health and safety have always been priorities in most companies, but working conditions now extend beyond basic issues of keeping the work area safe and clean. For example, as we learn more about ergonomic-related disorders such as carpal tunnel syndrome, employers have an even greater responsibility to incorporate health and safety factors into human resource plans. Other responsibilities include providing reasonable accommodations
Chapter 6
Human Resource Practices
to workers with disabilities or ensuring that male and female employees are protected from sexual harassment from fellow workers and others. Most companies provide many opportunities to contribute to the quality of working life. They can provide personal and career counseling, career development and employability services, recreational or cultural activities, daycare, special leave for family responsibilities or for community services, flexible work hours, outplace¬ ment services, and extended health care for retirees. Johnson & Johnson's Ethicon Endosurgery Division, in Blue Ash, Ohio, has a Wellness Center with exercise rooms and equipment to support employees in their manufacturing and R&D facility. Employees can use the center before or after working hours or during their breaks. In addition, those workers who are assembling products get regular, programmed "ergonomic" breaks every few hours, where they are required to do exercises designed to prevent repetitive motion injuries. All of these opportunities contribute to creating a more productive, safer, and more enjoyable in work environment. SAS Institute, Inc., consistently one of Fortune's "100 Best Companies to Work For," is a high-tech software development company based in Cary, North Carolina. SAS has a people-focused founder and CEO in the person of James Goodnight. Per¬ haps the most eye-opening policy of the firm is its mandated seven-hour workday. No "all-nighters" are expected of SAS employees. The multibillionaire Goodnight sets the example by leaving the office at 5 p.m., sharp. Many of the lavish employee perks at the sprawling corporate campus are family and lifestyle-oriented, from day¬ care centers, lactation rooms, a Montessori school, and a college prep private high school, to a 55,000-square-foot athletic facility, free massages, free car washes, and end-of-year bonuses. The payoff? SAS has about 4 percent turnover in an industry where 20 percent is the norm.60 Motivating Employees
Understanding human behavior and motivation are major elements of Deming's Pro¬ found Knowledge discussed in Chapter 3. Deming spoke of motivation as being pri¬ marily intrinsic (internal), and was suspicious of external forms of motivation, such as incentives and bonuses. Although thousands of studies have been performed over the years on human and animal subjects in attempts to define and refine the concept of motivation, it remains an extremely complex phenomenon that still is not fully understood. As managers in a TQ environment take on the roles of coaches and facil¬ itators, their skills in motivating employees become even more crucial. Saul W. Gellerman defined motivation as "the art of creating conditions that allow every There is no such thing as an unmoti¬ vated employee, hut the system one of us, warts and all, to get his work done within which people work can either at his own peak level of efficiency."61 A more seriously impede motivation or formal definition of motivation is an indi¬ enhance it. vidual's response to a felt need. Thus, some stim¬ ulus, or activating event, must spur the need to respond to that stimulus, generating the response itself. For example, an individual worker given the goal or quality task of achieving zero defects on the parts that he or she produces may feel a need to keep his or her job. Consequently, the worker is moti¬ vated by the stimulus of fear and responds by carefully producing parts to achieve the goal. Another less insecure worker may feel the need for approval of his or her work by peers or superiors and be motivated by the stimulus of pride. The worker then responds to that need and that stimulus by producing high-quality parts.
287
288
Part 2
Quality in High-Performance Organizations
Researchers have proposed many theories and models to describe how and why people are motivated. A theory is a way to describe, predict, and control what is observed in the world. Models graphically or symbolically show what a theory is saying in words. Often a model is so closely associated with a theory that the terms are used interchangeably. For example, Herzberg's Two-Factor theory describes two categories of factors, called "maintenance" and "motivational" factors. Maintenance factors are conditions that employees have come to expect, such as a safe working environment, a reasonable level of job security, supervision, and even adequate pay. Workers in a situation with these conditions will not be dissatisfied, but maintenance factors generally do not provide any motivation to work harder. Motivational factors, such as recognition, advancement, achievement, and the nature of the work itself are less tangible, but do motivate people to be more committed to and satisfied with their work. From Herzberg's theory arose the concept of job enrichment, described earlier in this chapter. With job enrichment, employees gain a sense of fulfillment (satisfac¬ tion) from completion of every cycle of a task. Acquiring cross-functional skills, working in teams, and increased empowerment are forms of job enrichment. Theories and models are often classified according to common themes. James L. Bowditch and Anthony F. Buono categorize motivation theories as content, process, and environmentally based theories.62 These theories are often studied in traditional management courses and are summarized in Table 6.6. In the behavioral sciences, as well as in the pure sciences, the originator of a theory is becoming more and more dif¬ ficult to determine because many researchers' ideas often overlap. Thus, the infor¬ mation in Table 6.6 is merely suggestive of one or more names that have been associated with the development of the theory. The Bonus Materials folder for this chapter on the CD-ROM contains detailed descriptions of some of these key theories, and we suggest that you review them. Motivation theories can be applied to support TQ in any organization. For example, Herzberg's theory suggests that ignoring maintenance factors such as supervision.
Table 6.6 A Classification of Motivation Theories Motivation Theory
Pioneer/Developer
Type of Theory
Abraham Maslow Frederick Herzberg Douglas McGregor David McClelland
Need Need/satisfaction Managerial expectations Acquired need
Victor H. Vroom Porter and Lawler Edward Locke Robert J. House
Expectancy Expectancy/reward Goal Goal
B. F. Skinner J. Stacy Adams A. Bandura Snyder and Williams
Reinforcement Equity
Content Theories Hierarchy of Needs Motivation and Maintenance Theory X-Y n-Ach, n-Aff, n-Pow Process Theories Preference-Expectancy Contingency Goal Setting Path-Goal Theory of Leadership Environmentally Based Theories Operant Conditioning Equity Social Learning/Seif-Efficacy
Social Learning/Seif-Efficacy
Chapter 6
Human Resource Practices 289
working conditions, salary, peer relations, status, and security will produce dissatisfac¬ tion and negatively impact the work environment, while enhancing the motivating fac¬ tors will produce a positive effect. Thus, understanding and applying the theories should result in more effective designs of work systems and the work environment. Performance Appraisal
Considerable truth can be found in the statement, "How one is evaluated determines how one performs." This reality can be dangerous. Analog Devices, a successful Massachusetts analog and digital equipment manufacturer, embraced TQ but found its stock price steadily declining. One of its key measures (on which managers were rewarded) was new product introduction time, with an objective of reducing it from 36 to 6 months. The product development team focused on this objective; as a result, engineers turned away from riskier new products and designed mundane deriva¬ tives of old products that no longer met customers' needs. The company subse¬ quently scrapped that goal.63 Performance appraisal is a process for evaluating and generating information about employees' effectiveness and efficiency at work.64 However, performance appraisal is an exceedingly difficult HRM activity. Organizations typically use per¬ formance appraisals for a number of reasons: to provide feedback to employees who can then recognize and build on their strengths and work on their weaknesses, to determine salary increases, to determine training needs, to identify people for pro¬ motion, and to deal with human resource legalities. As such, they can provide a paper trail to fight wrongful-discharge suits and act as a formal warning system to marginal employees.63 Many leading organizations use performance appraisal for changing corporate culture. Conventional appraisal processes typically involve setting objectives for a certain period of time (typically for the year ahead), either unilaterally or jointly by the man¬ ager with his or her subordinate. Objectives might focus on development of knowl¬ edge or skills, results such as output and productivity, or behavior. Objective setting is followed by a supervisory review of accomplishments, strengths and weaknesses, or personal characteristics of the subordinate related to the job at the end of the review period. Often, the form used for performance rating has 10 to 15 tangible and intangible categories, such as quantity of work, quality of work, works well with others, takes initiative, and so on, to be rated on a five- or seven-point scale from excellent' to "unsatisfactory" or "poor." The performance appraisal interview may be accompanied by announcements of raises, bonuses, or promotions. In some cases, company policy dictates a certain distribution of results, such as "no more than 10 percent of any department's employees may be rated as excellent" or "merit raises or bonuses will only be paid to employees who are rated as excellent or very good." Dissatisfaction with conventional performance appraisal systems is common among both managers, who are the appraisers, and workers, who are appraised. General Motors, for example, discovered that 90 percent of its people believed they were in the top 10 percent. How discouraging is it to be rated lower? Many managers are inclined to give higher ratings because of potential negative impacts. Numerous research studies over the past several decades have pointed out the problems and pitfalls of performance appraisals.66 Many legitimate objections can be made:67 • They tend to foster mediocrity and discourage risk taking. • They focus on short-term and measurable results, thereby discouraging long¬ term planning or thinking and ignoring important behaviors that are more dif¬ ficult to measure.
290
Part 2
Quality in High-Performance Organizations
• They focus on the individual and therefore tend to discourage or destroy team¬ work within and between departments. • The process is detection-oriented rather than prevention-oriented. • They are often unfair, since managers frequently do not possess observational accuracy. • They fail to distinguish between factors that are within the employees' control and system-determined factors that are beyond their control. Many companies use peer review, customer evaluations, and self-assessments as a part of the appraisal process. One approach that has been gaining increasing accep¬ tance and overcomes many of the objections cited earlier is called 360-degree feed¬ back.68 In an ideal 360-degree approach, a group of individuals who interact with the employee (or team) on a frequent basis participate in both the goal-setting process and the performance appraisal process. This group might include suppliers, clients, peers, internal customers, managers, and subordinates. The process involves twoway communication in which both parties discuss such needs as service levels, response times, accuracy of work and so on, which are often expressed as written ser¬ vice contracts. At the end of the performance period, selected representatives who participated in the goal setting evaluate how well the goals of the service contracts have been met, and provide feedback. The final performance appraisal consists of discussing an aggregation of the comments and ratings with the employee, and serves as a process for setting goals for the next period and for employee development. Because Performance appraisals are most the approach is new, little systematic research effective when they are based on the objectives that support the strategic has been performed on its effectiveness; how¬ directions of the organization, best ever, user feedback has been positive. practices, and continuous improve¬ In the spirit of Deming, many companies are ment. replacing performance evaluation altogether with personal planning and development sys¬ tems. Cadillac, for instance, replaced its traditional performance review with a per¬ sonnel development planning process in which managers meet with employees to set future expectations, identify training needs, provide coaching, and reward contin¬ uous improvement. Eastman Chemical Company eliminated employee labeling, improved the focus on individual development planning, and encouraged employee involvement and ownership. Granite Rock does not emphasize past performance, but sets professional development goals in conjunction with the company's needs. No stigma is attached to failure; the thrust of the process is to develop each indi¬ vidual to the fullest. Today, many leading organizations are focusing on identifying a small number of core competencies that are critical to the organization's success.69 These core compe¬ tencies are the behaviors, skills, and attributes every member is expected to have. They also use mastery descriptions, narratives of behavior that one who has mas¬ tered it would likely engage in. For example, a mastery description of Customer Focus might be: Dedicated to meeting the expectations and requirements of internal and external customers. Knows who every one of his/her customers is and can state what that individual's expectations are. Gets firsthand customer infor¬ mation and uses it for improvements in products and services. Speaks and acts with customers in mind. Takes the client's side in well-founded complaints. Is skilled at managing customer expectations. Establishes and maintains effective relationships with customers and gains their trust and
Chapter 6
Human Resource Practices
respect. Actively seeks customers' feedback on the quality of service he/she provides.
A behavioral frequency scale, in which appraisers indicate how frequently the appraisee does the things listed in the mastery descriptions (rarely, occasionally, fre¬ quently, or regularly, for example) is often used. This avoids numerical judgments of performance, defensive reactions, and provides a guide of what to do to improve. Measuring Employee Satisfaction and HRM Effectiveness
Measurement of employee satisfaction and HRM effectiveness is useful to assess the linkages with company strategy and to provide a foundation for improvement. In fact, research has suggested that organizations that use people measures as part of a balanced set of measures to manage the business see significantly higher return on investment and return on assets than those that don't. The same holds true for orga¬ nizations that say their employee surveys pro¬ vide valuable information to guide decision making. Nevertheless, few organizations have well-defined people measures or use them to predict key business outcomes.70 Both outcome and process measures provide data by which to assess HRM effec¬ tiveness. Outcome measures might include "hard" measures of cost savings, produc¬ tivity improvements, defect rate reduction, customer satisfaction improvements, cycle time reductions, and employee turnover, as well as "soft" measures of teamwork and management effectiveness, employee commitment, employee satisfaction, and empowerment. Typical process measures of success include the number of sug¬ gestions that employees make, the numbers of participants in project teams, and par¬ ticipation in educational programs. Team process effectiveness can be assessed by tracking the average time it takes to complete a process improvement project, and determining whether teams are getting better, smarter, and faster at performing improvements. Facilitators and program coordinators should also look for other indi¬ cators of success, such as improvements in team selection and planning processes, frequency of use of quality improvement tools by employees, employee under¬ standing of problem-solving approaches, and senior management involvement. Employee surveys can also help in providing this information. Questions in a typical survey might be grouped into such basic categories as quality of worklife, teamwork, communications, opportunities and training, facili¬ ties, leadership, compensation, benefits, and the company. Surveys might also address important team and individual behaviors, such as unity for a common pur¬ pose, listening effectively and acknowledging others' contributions, obtaining the participation of all members of the team, gathering and analyzing relevant data and information, sharing responsibility, using problem-solving processes and tools, and meeting company objectives for quality improvement. Many research-based and commercial survey instruments are available.71 Like the customer satisfaction surveys we discussed in Chapter 4, many employee surveys also seek feedback on the impor¬ tance of key issues. HR measures allow companies to pre¬ dict customer satisfaction, identify those issues that have the greatest impact on business performance, and allocate appropriate resources.
Employee surveys also help organizations better understand the "voice of the employee," particularly with regard to employee satisfaction, management policies, and their internal customers and suppliers. Such feedback helps organizations improve their human resource management practices. For example, Marlow Indus¬ tries uses a survey that addresses a broad variety of issues, including management
291
292
Part 2
Quality in High-Performance Organizations
support, the company's total quality system, organizational effectiveness, training, and continuous improvement. Table 6.7 shows most of the questions included in their survey. All responses are made on a five-item scale ranging from totally disagree to very much agree. Xerox produces its survey in 25 languages. Fifty-four questions are grouped into eight categories: Directions/communications. Valuing people. Trust, Learning, Feedback, Recognition, Participation/involvement, and Teamwork. Xerox compares results against similar companies such as Allied Signal, Honeywell, Sun Microsystems, Texas Instruments, and others. In evaluating results, trends and long-term results should be emphasized, and they should be communicated to employees. A good system should report results on a regular basis, perhaps monthly or quarterly, with a summary year-end report, using graphical aids wherever possible. Detailed reports should go to lower-level man¬ agers, showing results at their level. Summary reports should go to higher manage¬ ment levels. Specific action, such as training, changes in reward or recognition, or improvements to support employee well-being should be taken based on results. HRM in the Internet Age
Has the Internet Age changed what organizations need to know about HRM? The answer is surprisingly, no. Workers want to be treated with respect, have their basic needs addressed, understand the goals of their work, and have managers recognize their unique individual differences. They want to be given challenging, meaningful work in which they can experience pride of ownership, personal learning and growth, and be rewarded fairly and equitably when they perform. However, things have certainly changed in the fast-paced Internet Age. Managers who don't recognize these changes and keep up with trends in job design, motiva¬ tion, and leadership do so at their own peril. Take virtual teams, for example.72 These types of teams use a combination of Internet, e-mail, phone, fax, video conferencing, PC-to-PC connections, and shared computer screen technologies to get their jobs done. Virtual teams are often given a distinct Web site for posting charts, meeting minutes, statistics, and other shared documents. Virtual teaming requires special attention to communication, technology, sponsorship, and leadership issues. For example, the team leader needs to be able to tackle issues that he or she might not have encountered with traditional teams. One of the biggest disadvantages is the lack of experience members have working with one another. They are not aware of each other's work standards and cannot scrutinize these ethics as consistently as tradi¬ tional teams. This issue can be overcome by developing operating agreements by all team members, spelling out what they commit to do or not to do. Another factor is that communication is more complex because body language, voice inflection, and other communication cues are eliminated. Thus, virtual team members must be able to excel in relating their own ideas, and also understand the information others are trying to convey. Today's organizations are characterized by much less employee loyalty, much more organizational uncertainty (and paradoxically, much more individual opportu¬ nity), and much more dependence of the organization on its human capital than ever before. As one writer put it: "We are all temps (temporary workers)!" HRM practices in the Internet Age will increasingly require employers to take nontraditional approaches to attract and retain highly-skilled employees, including special perks, increased levels of responsibility early in the employees' careers, and understanding of the effects of changes brought about by new technologies and new realities of employee lifestyles in the twenty-first century.
Chapter 6
Human Resource Practices
293
Table 6.7 Employee Quality Survey—Marlow Industries Management Support 1. 2. 3. 4. 5. 6. 7.
The president is an active supporter of quality at Marlow Industries. Senior management (VPs) are active supporters of quality at Marlow Industries. My supervisor is an active supporter of quality at Marlow Industries. My supervisor is concerned more about the quality of my work than the quantity of my work My supervisor can help me to do my job better. My supervisor encourages good housekeeping efforts. I receive recognition for a top quality job done. Total Quality System
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.
Marlow Industries' Total Quality System is not a fad. It will be active long into the future. The Total Quality system has made an improvement in the performance of my work. The Total Quality system has made an improvement in my ability to do my job right the first time I understand the meaning of the Quality Policy. I believe in the meaning of the Quality Policy. I understand the meaning of the Quality Pledge. I believe in the meaning of the Quality Pledge. All departments within Marlow Industries support the Total Quality system. My co-workers support quality first. My co-workers believe in the Quality Pledge. My "supplier" co-worker treats me as his/her "customer" and meets my needs. I know who my internal "customer" is. I am able to meet the requirements of my internal customer. I believe that improving quality is the key to maintaining Marlow Industries' success. Organizational Effectiveness
1. I receive feedback that helps me perform my job better. 2. I am encouraged to stop and ask questions if something does not seem right. 3. There is a high level of quality in the products we ship to our external customers. 4. Marlow Industries provides reliable processes and equipment so that I can do my job right the first time. 5. I do not use defective materials. 6. I am provided proper procedures to do my job right. 7. My fellow workers have a high level of enthusiasm about Marlow Industries' quality. 8. I believe control charts will help us improve quality. 9. I believe Marlow Industries offers a high quality working environment. 10. I enjoy my job. Training 1. I have received training to be able to do my job right the first time. 2. I have received training on how to determine if the work I do conforms to Marlow Industries' workman¬ ship standards, and other requirements of the customer. 3. I receive adequate safety training so that I am aware of the safety and health requirements of my job. 4. My supervisor has received adequate training to be able to do his/her job right the first time. 5. My co-worker has received adequate training to be able to do his/her job right the first time. 6. I have received ongoing training. 7. The training I have received has been very helpful to me in my job. Job Satisfaction and Morale 1. I have a high level of personal job satisfaction. 2. My morale is high. 3. The morale of my work group is high. Involvement 1. I feel involved at Marlow Industries. 2. I would like to be more involved at Marlow Industries.
Source: Courtesy Marlow Industries.
294
Part 2
Quality in High-Performance Organizations
HUMAN RESOURCE FOCUS IN THE BALDRIGE CRITERIA, ISO 9000, AND SIX SIGMA
Category 5 of the 2003 Malcolm Baldrige-National Quality Award Criteria for Perfor¬ mance Excellence is Human Resource Focus. This category examines how an organiza¬ tion's work systems and human resource practices lead to performance excellence and align with strategic objectives and action plans. Item 5.1, Work Systems, focuses on how work and jobs are organized to promote cooperation, initiative, empower¬ ment, innovation, and organizational culture, how they capitalize on diversity within the organization, and how communication and skill sharing are accomplished across work units, jobs, and locations. It also examines the performance management system, and how compensation and recognition reinforce high-performance work. Finally, it addresses hiring and career progression approaches, including effective succession planning for senior leadership. Item 5.2, Employee Learning and Motivation, focuses on how employee education, training, and career development support the achievement of objectives and build employee knowledge, skills, and capabilities. It examines how key organizational needs, such as performance measurement and improvement, technological change, ethical business practices, leadership development, and safety are addressed through education and training, how needs are determined, how education and training are delivered and reinforced on the job, and how the effectiveness of education and training are evaluated. This item also addresses how an organization motivates employees to develop and utilize their full potential, and attain job- and careerrelated development and learning objectives. Item 5.3, Employee Well-Being and Satisfaction, examines how an organization ensures a safe and healthful work environment and support climate that contribute to the well-being, satisfaction, and motivation of all employees. It includes identi¬ fying appropriate measures and targets for key workplace factors so that status and progress can be tracked, and how preparation for emergencies or disasters is addressed. The criteria also asks how the organization determines the key factors that affect employee well-being, satisfaction, and motivation; how services, benefits, and policies support employees; how well-being, satisfaction, and motivation are assessed and measured; and how assessment findings are used to identify priorities for improving the work environment and employee support climate. The focus of human resources in ISO 9000:2000 revolves primarily around training and the work environment, but does not address the subject as comprehen¬ sively as Baldrige does. The standards require that "Personnel performing work affecting product quality shall be competent on the basis of appropriate education, training, skills, and experience." The standards further require that organizations determine the level of competence that employees need, provide training or other means to ensure competency, evaluate the effectiveness of training or other actions taken, ensure that employees are aware of how their work contributes to quality objectives, and maintain appropriate records of education, training, and experience. The standards also address the work environment from the standpoint of providing buildings, workspace, utilities, equipment, and supporting services needed to achieve conformity to product requirements, as well as determining and managing the work environment, including safety, ergonomics, and environmental factors. Human resource focus is essential to Six Sigma. We discussed the role of project teams in Six Sigma earlier in this chapter. One quality professional noted that "Six Sigma actually owes its success to all the quality efforts that have come before it, and teams are an integral part of Six Sigma implementation."73 In addition to teams.
Chapter 6
Human Resource Practices
295
selecting the right people to serve on teams, training and skill development, and reward and recognition approaches to drive behavior are vital to Six Sigma efforts. Six Sigma efforts often result in significant change recommendations to the organiza¬ tion; work processes change and employees need to do things differently. Under¬ standing how changes affect people is a necessary issue that organizations must address after Six Sigma projects are completed; project champions, in particular, need to apply the principles discussed in this chapter to their organizations.
Quality in Practice
TD Industries74 TD Industries is an employee-owned firm that pro¬ vides mechanical, refrigeration, electrical, plumbing, building controls, and energy services to customers in Texas and the Southwest. Head¬ quartered in Dallas, the company expanded rapidly in the 1990s, due to the tremendous growth rate of the area. In the early 1990s the company employed about 600 people and had revenues of approximately $75 million per year. By the end of 1998, employment had grown to 1,050 employees, with revenues of $182 million. It is interesting to note that the company added 98 jobs over the 1997-98 timeframe, but had 2,200 applications for those jobs! What made these jobs so attractive in an industry where the norm is hard physical work, highly cyclical demand, and typically a high employee turnover rate? To help answer this ques¬ tion, consider the company's vision: We are committed to providing out¬ standing career opportunities by exceeding our customers' expectations through continuous aggressive improvement.
Note the focus on its people as reflected in the goal of providing outstanding career opportunities. A list of some of the awards it has won attests to its ability to achieve this vision: • 1998, Texas Quality Award (based on the Malcolm Baldrige criteria) • 1997, No. 5 on Fortune Magazine's list of "The 100 Best Companies to Work for in America" [No. 2 in 1998 and No. 4 in 1999] • 1996, National Member of the Year Associ¬ ated Builders and Contractors • 1996, Jack Lowe, CEO—Crystal Achievement Award National Association of Women in Construction
• 1996, National Carrier Distinguished Dealer Award • 1995, Commercial Contractor of the Year, Contracting Business • 1995,1996,1997, United Way Pacesetter/Elite Company In testifying to the members of the U.S. House of Representatives Committee on Education and the Workforce (http://edworkforce.house.gov/) on May 20, 1998, regarding "The American Worker at a Crossroads Project," Ben Houston, President of TD Industries shared the company philosophy: We have humbly accepted the above awards on behalf of all Partners of TD Industries because we believe our culture is founded upon having fun while we accomplish the following: Trust, Servant Leadership, Quality, Sharing. To have fun, every meeting agenda item begins with Item #1 — HUMOR. Our entire culture is based on a trusting relationship between all of our stakeholders—our clients, our communities, our partners, and our suppliers. We believe in the value of the individual, and attempt, in our culture, to recognize this value. TD Industries believes in Robert Greenleaf's philosophy as indicated in his book entitled, The Servant Leader. We believe that all individuals within TD Industries, regardless of their jobs, are leaders and managers of their own jobs. We also believe that in order to lead, one must first serve those that are to be led. We believe everything we do must strive towards meeting our stakeholders needs —our clients, our communities, our partners, and our suppliers. In order to ensure quality, teamwork among all TD
Part 2
296 Partners, our clients, and suppliers is essential. Most opportunities for improve¬ ment are accomplished through quality work teams. We avoid the term "buy in" because it implies someone has the cor¬ rect answer and must sell it to others. We prefer to have the best solution created by equal input from all participants.
In order to ensure improvement, TD Industries measures many items on an ongoing basis and the information is available to all TD Partners, but especially the work group that can affect the improvement. Items measured include the following: 1. Partner Satisfaction: An annual confidential questionnaire is compared with national groups, benchmarked against TD's own per¬ formance year by year to ensure that improvement is being made in all areas. 2. Supervisors Survey: Supervisors are confiden¬ tially rated annually by the partners who work with them, and the reports are given to the supervisors to be reviewed with their supervisor within three weeks. 3. Customer Surveys: Every area of customer con¬ tact is surveyed on a regular basis depending on the type of business. A goal of 9 out of 10 has been set, and all scores below 7 are reviewed with the client and the responsible manager and reviewed for lessons learned. 4. One With Ones: One with One reviews are required yearly with each partner and his or her supervisor and not at the same time as pay evaluations. 5. Pay Evaluations: Pay evaluations are done twice annually with training records and career paths being a portion of the review. 6. Productivity Reports: Sent to the partners doing the work. 7. Safety Reports: Sent to the partners doing the work. 8. Continuous Improvement: All partners, regard¬ less of their position in the company, are asked to obtain a minimum of 32 hours of training in order to sharpen the skills of the individual, and assist in ensuring the contin¬
Quality in High-Performance Organizations
uous improvement of the company's greatest resource—its employees. In 'addition, formal training for leaders includes safety orientation; mentors; 90-day orien¬ tation; sixth-month benefit orientation; one-year TD Opportunities; second year Quality and years 3 through 6 Leadership Training on Diversity, Team Work, Leadership, and 7 Habits. Associated Builders and Contractors Wheels of Learning craft courses are offered as well as technical and man¬ agement courses. TD Industries provides many benefits to its employees: medical, dental, group term life, and long-term disability insurance; employee stock ownership plan (ESOP)/401(k)—30 percent of the profits go to partners in the ESOP and 401 (k) plans; and other benefits such as paid personal time, holidays, sick pay, job injury pay, jury duty, wellness program, funeral pay, continuing educa¬ tion program (fully paid for by TD Industries through which many have obtained degrees at night), prescription safety glasses, and partner grant requests with up to $100 for a community service in which the partner works. Partners are encouraged to participate in many associations such as Associated Builders and Contractors and United Way in order to give back to the industry and community. Financial incentives are paid to virtually every partner based on improvements in the partner's own operation if partner satisfaction, supervisor scores and customer satisfaction are acceptable. Thirty percent of all pretax profits are distributed to the partners in the ESOP and 401(k). As Houston noted, "These items of SHARING dose the loop of TRUST due to the sharing that TD Industries practices with all Partners." Key Issues for Discussion
1. Explain how human resource activities at TD Industries work toward achieving the com¬ pany's vision statement. 2. How do HRM processes at TD Industries support the fundamental principles of TQ: customer focus, participation and teamwork, and continuous improvement?
Chapter 6
Human Resource Practices
29 7
Quality in Practice L.L. Bean75 L.L. Bean Co., a sporting goods and apparel mail¬ order distributor and retailer headquartered in Freeport, Maine, has been known for quality and a focus on the customer since it was founded in 1912 by Leon Leonwood Bean, a Maine outdoorsman. "L.L.," as he was called, grew tired of coming home with wet, sore feet from the heavy leather woodsman's boots of his day. He invented a new kind of boot that combined lightweight leather tops with waterproof rubber bottoms, incorpo¬ rating the best features of both materials. The practical advantages of his new L.L. Bean boots were readily apparent, and he soon sold 100 pairs to fellow sportsmen through the mail. Unfortu¬ nately, 90 pairs were sent back when the stitching gave way. But L.L. was true to his word, refunded his customers' money and started over with an improved boot. L.L. Bean operated his business based on the following belief: "Sell good merchandise at a rea¬ sonable profit, treat your customers like human beings, and they will always come back for more." The company now sells more than 16,000 outdoor products, and even has eight stores in Japan. The Freeport flagship store is one of the most popular tourist destinations in Maine, receiving more than 3.5 million visitors a year. The company also has 12 factory outlet stores: six in New England, five in the mid-Atlantic states, and one in Oregon. L.L.'s prod¬ ucts were originally sold only through the mail, but so many people dropped by his Freeport workshop to purchase items that he opened a showroom in 1917. Over the years, the company kept growing as L.L. added casual and sports apparel, gear and other footwear to his line. By 1951, people were dropping by day and night on their way to hunt and fish in Maine, and L.L. announced he had "thrown away the keys," deciding to keep the retail store open continuously. It has been open to the public 24 hours a day, 365 days a year, ever since. Legendary stories are told of commitment to quality and service in its catalogs and inside and around the company. One such story was told of customer service representative who was informed of a late shipment, loaded a canoe on his car, and drove from Freeport to New York City (a
distance of almost 300 miles, or 475 km) so that the customer could go hunting the next morning. Bean backs up its reputation for quality and cus¬ tomer focus with a "Guarantee of 100% Satisfac¬ tion" on all products. For 20 years. Bean had enjoyed 20 percent annual growth in sales. In the 1990s, the firm was caught in a major industry downturn, with rising costs and shrinking demand. Bean launched a TQ and HR effort in the mid-1990s. The focus on quality was led by an 85-person total quality and human resources (TQHR) unit, composed of HR generalists, some HR specialists, and process improvement consultants. Bean is a people-focused organization, so mas¬ sive layoffs were not seen as a viable cost-cutting option. Bob Peixotto, vice president (total quality and human resources) stated, "People are the solu¬ tion, not the problem." Traditionally, the company had paid an average of 15 percent of salary as a profit-sharing bonus, each year. During the 1990s, bonuses were unpredictable and sporadic. Limited voluntary retirements and staff reductions were carried out, but TQHR saw that the solution was to "work smarter, not harder." As a direct marketing and retail business, large-scale complexity and dependence on worldclass business logistics put people, total quality, and customer service at the heart of the company. To save time and meet customer needs, continual process review and improvement were essential. L.L. Bean began to rethink each of these elements. L.L. Bean already had an intense customer focus, so the two key drivers of the total quality effort were people (their involvement in the busi¬ ness) and processes (business process manage¬ ment and improvement). Anew people-process model helped to integrate TQ into the HR depart¬ ment. Total quality was defined by Bean employees as "managing an enterprise to maxi¬ mize customer satisfaction in the most efficient and effective way by totally involving people in improving the way work is done." TQHR's mis¬ sion was to support the company TQ mission by acting as a catalyst for total quality and superior performance.
298 For a retail organization like L.L. Bean, quality is not a production line issue. As Peixotto said, "It happens every time a customer representative receives a call in the telephone center. That interac¬ tion is where quality really happens, so a third goal was to change the infrastructure to support such customer interaction. This eventually meant a redesigned TQHR department." To accomplish their TQ goals, departmental leadership decided to focus its own efforts on five operational themes: 1. Servicing line managers to meet employee needs, role revision for managers and man¬ agement learning. 2. Practicing what it preached—with its own performance improvement and management being one example. 3. Acting as one unified department with a portfolio of products and services. 4 Regarding itself as a business, with regular feedback from internal customers being crit¬ ical evidence of performance. 5. Continually reviewing and rethinking the TQHR process. When TQHR reviewed its work to identify the types of business in which it should be involved, almost 50 different kinds emerged, including busi¬ ness process improvement, publishing, facilitation, management consulting, report management, legal care management, physical therapy, learning center operations, and career counseling. Next, the department devised a portfolio of TQHR products or services for internal customers for which resources were needed. The final list of 109 included career assessment, ergonomic work¬ station design, quality assessments, and change management counseling. Six core processes were identified for delivering these products and ser¬ vices to other parts of the business, including recruiting and orienting people, developing people, separating people, organizational develop¬ ment planning, product and service development, and delivering products and services. Peixotto explains, "These changes helped us to create a new TQHR paradigm, where depart¬ mental staff worked directly in customer areas and therefore at the front end of the business and total quality. Our role here was to act as consultants, not order takers for HR products or services." As a result of the rethinking, the new TQHR organiza¬
Part 2
Quality in High-Performance Organizations
tion was divided into two sections—the resource center and service teams that work in internal-cus¬ tomer areas across all L.L. Bean's operations. The resource center acts as a strategic HR thinktank, deciding what TQHR should be offering and examining the strategic needs of internal customers in relation to employee performance, productivity, and continuous improvement. The service teams are business partners, delivering products and ser¬ vices directly for internal customers, working away from their own operational area. To accomplish work across departmental units, TQHR teams called "pods," were also introduced. These short-life teams are rapidly and flexibly deployed process-improvement consultants—"the catalysts for change who go after costs," in Peixotto's words. These pods are expert teams that address needs such as training and development, recruitment and compensation, and can assume both strategic and operational roles for their work. For example, the compensation pod is investi¬ gating new, major TQHR approaches such as team-based rewards and recognition. Broadly skilled business people are assigned to work in TQHR. Their skills include both traditional HR competencies, such as training and recruit¬ ment, and specific process improvement skills such as systems, information technology, and industrial engineering. Thus, they can be deployed across the resource center, pods, or service teams. Peixotto's unit also does not hesitate to use talent from elsewhere within the organization. For example, any L.L. Bean employee with appro¬ priate skills is invited to attend the annual training courses in facilitation and process improvement. In return, these "outside" individuals guarantee TQHR 52 days of personal input a year to support its activities. He says, "They act as skilled per¬ sonnel and change or process improvement mis¬ sionaries. Broad and interrelated consultancy skills will always be TQHR's critical capability. But, as internal consultants, we never tell any part of the organization what it should be doing: our job is to listen first and then recommend action or tools for getting things done effectively." Peixotto uses the TQHR unit as a test labora¬ tory for the organization's HR tools or techniques to support its work in the continuous improve¬ ment process. It requires that TQHR anticipate internal-customer requirements or how their needs might change with new business priorities.
Chapter 6
Human Resource Practices
299 Then, they test out the tools and techniques that range from the basics—pay, incentives, training, competencies, performance measures, employee relations and so on—to more complex approaches, including job redesign and process improvement. They may be applied in one of two scenarios: (1) as tested templates for managers to use to save on their own time and resources and avoid rein¬ venting the wheel; or (2) for service teams, where they have been invited to help in a function or business area. "The invitation is important," explains Peixotto, "because our role is to act as facilitators and enablers." TQHR also benchmarks itself with any organization that has proven best practices. Recently, performance management, leadership development, idea generation systems, compensation issues and workplace safety have been benchmarked. TQHR departmental productivity savings have reduced operating expenses by $500,000. Addi¬ tional direct savings from TQHR-driven initiatives in one year were $5.6 million from business process improvement projects, $2.75 million saved in health costs because of better working practices, and $2.1 million through workforce planning. With a renewed focus on high customer service and quality standards, Peixotto's reengineered process is receiving high marks for internal cus¬ tomer satisfaction. The highest rated attributes of the TQHR unit are customer responsiveness, "easy to do business with," accessibility and flexibility. Previously, these qualities were among those the original HR department performed less well, which contributed to a loss of organizational credi¬ bility. Peixotto now feels that the redesigned and repositioned TQHR unit is well placed (and expert enough) to work in partnership with other sup¬ port areas in the company. They are now begin¬
jjgfl
ning to have a substantial impact on quality prob¬ lems through the use of process improvement pro¬ jects and, where appropriate, by reengineering processes or by specific departmental activities. TQHR is now a kind of "amoebic unit," which Peixotto describes as teams separating and forming as required, continuously learning and operating as if every day is a new pilot project. He comments, "Change is the organization's constant. TQHR s job is to increase its own effectiveness first, to help people to understand their processes and to cope with the level of change we're cre¬ ating. The big name of the game for the late 1990s will be managing change." Bob Peixotto says that L.L. Bean's traditional HR department, the forerunner of the total quality and human resources unit, was viewed as "the necessary organizational stepchild and extra busi¬ ness cost." The reengineered, customer-focused department, in contrast, is described as "easy to do business with," accessible and flexible. If the name of the game for the turn of the century is to be managing change, the TQHR unit at L.L. Bean will surely be at the forefront. Key Issues for Discussion 1. Discuss L.L. Bean's approach to blending human resource management with total quality. What lessons can be applied to other organizations? 2. How has the TQHR Department modeled the characteristics of teams given in Scholtes' list of successful team characteristics? 3. Because L.L. Bean is primarily a service firm, what special challenges does it face as it attempts to emphasize quality to its employees?
Review Questions
1. Discuss the impact of the Taylor system on quality, productivity, and human resource management. How has TQ changed business thinking about Taylor? 2. Define human resource management. Contrast it with the traditional role of per¬ sonnel management. 3. Contrast traditional HRM approaches with those required in a TQ environment. 4. Summarize the leading HRM practices encountered in TQ organizations. 5. What is a team? Define the major types of teams found in organizations today.
Part 2
300
Quality in High-Performance Organizations
6. Contrast the differences between quality circles and self-managed teams. What are the key characteristics of self-managed teams not found in quality circles? 7. What roles must be fulfilled within the structure of teams? What steps can team leaders take to coach less experienced team members to enhance team perfor¬ mance? 8. Discuss the four phases that teams typically go through during their life cycle. 9. Explain the important issues an organization must consider in developing suc¬ cessful teams. 10. How are the roles in Six Sigma teams similar to and different from traditional project teams? 11. What is high-performance work? What types of HR practices contribute to a highperformance work environment? 12. Explain the difference between work design and job design. How does the Hackman and Oldham model enhance understanding of how job design affects motivation, satisfaction, and organizational effectiveness? 13. What is employee involvement? Discuss some of the early developments of El approaches. What are the advantages of El over traditional management prac¬ tices? 14. How can managers overcome resistance to El initiatives? 15. What is empowerment? Discuss the changes that empowerment brings to orga¬ nizations. 16. Discuss the role of training and education in supporting total quality. 17. What types of compensation practices support total quality? 18. What are the key practices that lead to effective recognition and reward approaches? 19. What issues must organizations consider with respect to health, safety, and employee well-being in the work environment? 20. In a TQ-based organization, what is the role of recruitment and career develop¬ ment? What challenges does TQ pose in these areas? 21. Define the term motivation. Why is motivation critical in a TQ environment? 22. Briefly summarize traditional performance appraisal processes. Within a TQ perspective, what objections have been raised concerning these processes? What steps can be taken to make performance appraisal more consistent with TQ principles? 23. What is 360-degree feedback? How does it differ from traditional performance appraisal approaches? How does it address the major criticisms of traditional performance appraisal processes and support TQ efforts? 24. Why is it important to measure employee satisfaction and HRM effectiveness? Describe some common approaches. 25. Summarize the HRM issues addressed in the Malcolm Baldrige National Quality Award criteria, ISO 9000:2000, and the Six Sigma philosophy.
Discussion Questions 1. What is your reaction to the quote about Toyota in the opening paragraph of the chapter? Would such an observation be true of most other organizations? Is it really true that competitors cannot copy the human resources of an organiza¬ tion? Why or why not?
Chapter 6
Human Resource Practices
2. Peter Drucker, arguably the most respected and influential writer on manage¬ ment, observed: Whatever his limitations and shortcomings-and he had many-no other American, not even Henry Ford (1863-1947), had anything like Taylor's impact. "Scientific Management: (and its successor, "Inudstrial Engineering") is the one American philosophy that has swept the world-more so even than the Constitution and the Federalist Papers. In the last century there has been only one worldwide philosophy that could compete with Taylor's: Marxism. And in the end Taylor has triumphed over Marx.76 Comment on Drucker's observations about the Taylor system. Do you agree with his statement about Taylor versus Marx? Why or why not? 3. What can an organization do about individuals who "aren't good with num¬ bers" if they have a policy that they become Green Belts, and later. Black Belts, as a prerequisite for promotion to higher levels of management? 4. How can a fraternity or student organization use leading HRM practices of companies to develop its own strategic HRM plans? If you are involved in such an organization, develop a strategic HRM plan that supports total quality. 5. Think of a job you have had. Apply the Hackman and Oldham model to eval¬ uate how the job design impacted your motivation and satisfaction, as well as organizational effectiveness. 6. Recently, new "employee performance software" has been developed to track individual output. For example, British Airways uses it to ensure that customer service reps' time in the break room or on personal calls doesn't count on the clock. The technology can keep track so that extra incentive dollars are eventually directed in to the paychecks of those whose digital records merit them. It can also help managers understand how to assemble the most effective teams or who to lay off.77 Discuss the implications of such technology from a TQ perspective. 7. Cite some examples of empowerment or lack of empowerment from your own experiences. 8. How might the concept of empowerment be employed in a classroom? 9. Undoubtedly you have received a recorded message prior to a call to company that says something like "For quality purposes, this call may be recorded." What do you think the real purpose of such an approach is? Is it to improve quality or to monitor poorly trained employees or catch them deviating from company scripts? Would an empowered organization need to use this method? 10. Many companies today seek the best available applicants and train them in TQ piinciples. What implications does this practice have for designing college cur¬ ricula and choosing elective courses in a given program? 11. How might a jazz quartet be viewed as a metaphor for a team in a business sit¬ uation? 12. Students in early grades often receive many kinds of recognition: stickers, candy, and so on, for good work. As we discussed, similar forms of recognition are common in the workplace. Yet little daily recognition is given at the high school and college level. Discuss possible reasons for this difference and design a recognition program that might be appropriate in your class. 13. Consider the statement, "How one is evaluated determines how one performs." What does this notion mean for your classes? Would your performance change if grades were abolished (as Deming strongly advocated)? 14. Discuss the controversy over performance appraisal. Do you agree with Deming's approach, or do you take the more traditional viewpoint toward per¬ formance review? Why?
301
302
Part 2
Quality in High-Performance Organizations
15. Most colleges and universities use a course/instructor evaluation system. If your school has one, how is it used? Does it support continuous improvement or is it used strictly for performance appraisal? How might the evaluation instrument or process be modified to better reflect TQ principles? 16. Jack Welch, former CEO of General Electric, stated his passion for making people GE's core competency. He used a system in which executives in the bottom 10 percent of a forced performance ranking were eliminated. What do you think of this approach? How does it fit with a TQ philosophy? How would you respond to someone who says, "I think all my people are pretty good. If I fire the bottom 10 percent, that would just give me a new bottom 10 percent. Where does it end?" Read the paper at http://www.worklab-consulting.com (click on IQ) for some perspectives. 17. The training strategy that Xerox used is summarized as follows: a. The training is uniform—common tools and processes are taught across all of Xerox, to all employees, creating a "common language within Xerox" that fosters cohesive team functioning. b. Training is conducted in family groups, with all members starting and fin¬ ishing training at the same time to facilitate the change process. c. Training starts at the top of the organization with the CEO and cascades downward to all employees.78 What advantages does such strategy have? Do you see any possible disadvan¬ tages? Would this approach work in any business? 18. Discuss the conditions under which team incentives, gainsharing, and "pay for increased skills" reward systems may work. When is it a poor idea to install such systems? 19. A restaurant owner noted that "waitstaff skills are very trainable; human-being skills are not. I can train anyone to be knowledgeable about our wine list or how to clear a table properly. But I cannot teach people to care about how their actions affect others." Do you agree with the statement? 20. What motivates you to study and perform in the classroom? How do motivation theories apply to you personally? Discuss how these theories might lead to new ways of teaching and learning. 21. When simple theories such as those of Maslow, Herzberg, and McGregor explain motivation, why does the search continue for more complex ones or for ones that integrate several different theories, such as Porter and Lawler's theory? What implications do they have for quality? (Read the Bonus Materials on Theories of Motivation first.) 22. Suppose that someone told you that she was just promoted to manager of a department with several "star" employees who are constantly getting new job offers. Aside from compensation issues (assume they are well-paid), what might you suggest as means of ensuring that these employees remain loyal to the company? Labor relations between unions and management can make it difficult to establish TQ-oriented HRM practices within organizations. 23. The National Labor Relations Board (NLRB) ruled on two cases in 1993 and 1994 that complicate a company's determination of how far it can go legally to set up and use employee participation programs (EPPs) to make improvements in the workplace. The two cases involved a small, nonunion company, Electromation, and a large company, DuPont. These case decisions by a five-person board were based on interpretations of the 58-year-old National Labor Relations Act (NLRA, or Wagner Act) that prohibits unfair labor practices. The rulings are found in the
Chapter 6
Human Resource Practices
303 NLRB proceedings as Electromation vs. International Brotherhood of Teamsters (309 NLRB-No. 163), and E. I. duPont de Nemours and Company vs. Chemical Workers Association, Inc. (311 NLRB-No. 88). In the Electromation case, the nonunion com¬ pany s management set up five employee action committees to deal with poli¬ cies concerning absenteeism, smoking, communications, pay for premium positions and attendance bonuses. In DuPont's case, management unilaterally (without bargaining with the union) changed the composition of safety and fit¬ ness committees to include nonmanagerial employees (where the committees ad previously been composed only of management) at a unionized New Jersey plant. Stated briefly, the cases specified that "employer-dominated labor organi¬ zations" are prohibited. In both cases, the employee teams/committees were ruled to be "labor organizations" and to be "management dominated." Discuss the implications of these cases, particularly in the context of TQ. You might wish to conduct further research on these cases.
Etc 1. Briefly review the history of HRM. Conduct a thorough literature search of one of t e branches of HRM and relate it to current quality management issues. Search some current business periodicals (e.g.. Fortune, Business Week) for arti¬ cles dealing with HRM issues. Explain how they relate to the material in this chapter. Are any new approaches emerging? 2. Interview managers at a local organization about their HRM practices, focusing on work and job design issues. Report on your perceptions of how well their practices support a high-performance workplace. 3. Survey local companies to determine if and how they use suggestion systems. What levels of participation do they have? Are suggestions tied to rewards and recognitions? 4. Investigate the extent of team participation at some local companies. What kinds of teams do you find? Do managers believe these teams are effective? 5. Find a small to medium-sized company that is using Six Sigma teams. Have they changed the GE/Motorola model in the way that they train and use team leaders and resource people (Green, Black, and Master Black Belts)? Are they using those roles for management development purposes? 6. Survey several managers in one or two companies on the topic of motivation for quality. Try to find managers at each of the following levels to interview: a. Quality control/assurance b. Manufacturing or industrial engineering c. Upper-level management d. First-line supervision e. Line employees (perhaps a union steward or officer) 7. Research the impacts of the Internet Age on human resource practices in an actual firm. One possible approach would be to interview an HR manager at a company that is changing from a bricks-and-mortar to an Internet-based orga¬ nization (such as a telephone company that is shifting to a broad-based com¬ munications firm). Another approach might be to visit the Web sites of several firms, examine HR practices that may be described , and compare and contrast your findings.
304
Part 2
Quality in High-Performance Organizations
HL^ I. The Hopeful Telecommuter Jennifer Smith was pregnant, and she was happy about it. She and her husband, Jim, had been plan¬ ning to start a family for some time. However, she was concerned about her job as a Northeast Zone supply chain manager for health and beauty prod¬ ucts for Big Bear Stores. Big Bear was a large, multibillion dollar foodstore chain that had stores in 47 states. It was a conventionally organized retailer divided into three geographic regions (Atlantic, Mid-American, and Western) with 12 zones (4 per region). Zone supply chain managers, such as Jennifer, were the link between the store managers and their product-line suppliers. Jennifer had been ranked number one in customer and in supplier satisfaction surveys for health and beauty product lines for the last two years. She knew that she was eligible for six months of maternity leave under the federal Family Leave Act, and that the com¬ pany would have to provide a job for her upon her return. What she didn't like was the thought that they did not have to, and probably would not, give her the same job that she was now holding so well. Jennifer had talked with Jim, at length, about what to do. They agreed that she should approach her regional manager, Sarah Strong, the Zone VP, about the possibility of "telecommuting" to her job after the baby came. Jennifer thought that she could do 85-90 percent of the job at home on her own schedule. A large part of her job consisted of verbal and fax contacts with store managers and suppliers, as well as extensive use of a computer for manipulating databases, preparing spread¬ sheet reports, and sending and responding to email. The other 10-15 percent of the time, when she had to be in the office for face-to-face meetings or had to take brief trips, her parents and Jim could keep the baby and cover for her at home. When Jennifer approached Sarah Strong, she
was interested, but would not commit herself to supporting Jennifer's request to telecommute. She said that the company had never done that before, and it might pose a number of difficulties. She did say that she would take her request forward to the two V.P.'s who could approve or disapprove it. Both senior managers would have to approve Jen¬ nifer's request, however. Sarah asked Jennifer to prepare some "talking points" concerning the ben¬ efits versus the limitations of the arrangement that she could present to the vice president of human resources, and the senior vice president of opera¬ tions, Sarah's manager. She also asked her to pre¬ pare a cost estimate, in consultation with the Zone information systems manager. The following was what Jennifer prepared for the estimated costs: Laptop computer and docking station Setup DSL dedicated phone line Fax machine Computer desk and chair Telephone line charges (6 months)
$3,500.00 250.00 250.00 375.00 240.00
Total
$4,615.00 «
Discussion Questions
1. You are Jennifer. What "talking points" would you prepare to support your case? Include both the strengths and limitations of telecommuting. Keep in mind the needs of your "customers," the human resources VP, as well as Sarah, and the VP of operations. 2. What issues do you think that the VP of human resources might raise? What issues do you think the senior VP of operations might raise? 3. How does your answer demonstrate the principles of empowerment? How might it fit the components of the Hackman-Oldham Job Characteristics model?
Chapter 6
Human Resource Practices
305
II. Crystal Silicon, Inc.
Crystal Silicon, Inc. (CSI), manufactures ultra-pure silicon wafers for the semiconductor industry located in the midwestern United States.7" The plant employs approximately 600 people. The hourly workforce is nonunion, and most are hired without much experience in the industry and are trained locally, because it is difficult to entice expe¬ rienced and highly paid employees from the West Coast to move to the Midwest location. Electronics engineer Vernon L. Essai (pro¬ nounced S-eye) founded CSI in the early 1980s. Essai had worked for Texas Instruments and Intel for 10 years. He learned the value of innovation and was able to stay ahead of the technology curve while running a small company by being agile and stressing high quality of products. He was finding it increasingly hard to attract innova¬ tive product designers in what had become a worldwide marketplace for people with talent. CSI's management structure is pretty tradi¬ tional, with a highly educated, technically trained design department, and the functional depart¬ ments of marketing, production, administration, and finance, each presided over by one of the five vice presidents. Direct supervisors conduct performance appraisals of all employees, and identify key objectives for personal development that are reviewed by the department managers. A prin¬ cipal focus of this process is to identify the things that employees will need to do for promotions and career development. Supervisors are trained to look for high potential in workers and continually try to motivate them to seek career development opportunities. These appraisals are supplemented by formal yearly performance evaluations. Hourly employees are rated on safety, quality, production output, dependability, initiative, judgment, team¬ work, and attendance. Salaried employees are rated on safety, knowledge, planning and organi¬ zation, execution and results, initiative, creativity, teamwork, and communication. These ratings directly affect their yearly merit increase. The HR department trains all new managers and supervi¬ sors on how to conduct performance appraisals. During the yearly planning process, depart¬ ment managers work with senior managers to determine training needs. Senior management
then meets with the HR director to determine what the budget can support. Rewards are given for perfect attendance, and all employees participate in a financial incentive program, which is paid out quarterly. Financial bonuses are based upon results of on-time delivery, customer complaints, process yield, and lost work-day accidents. Bucking an industry trend, stock options have not been used to reward either senior managers for exceptional results, or design team members, who have "scored home runs" in designing innovative and profitable prod¬ ucts. Essai was asked about it by one of the vice presidents, and he replied, "Why should I give away the company to people who are just doing their jobs? If we see that they are well paid for their work, why would they want to work harder for some gamble that our stock will go up in the future?" CSI uses an employee survey to determine sources of dissatisfaction and opportunities for improvement. Survey questions measure behav¬ ioral impacts on productivity, profitability, cus¬ tomer satisfaction, and turnover. The data are segmented by type of employee (i.e., salaried vs. hourly) as well as by job type and department. Despite CSTs careful attention to systematic employee appraisal, fair compensation, and finan¬ cial rewards for accomplishment of goals, and in the face of a technology downturn, turnover recently increased substantially. The employee survey showed pockets of dissatisfaction among both hourly and salaried workers related to oppor¬ tunities for challenge on the job and growth in knowledge of how to perform jobs better. Discussion Questions
1. How does CSTs approach to performance appraisal support high performance work? Discuss the differences in evaluating hourly and salaried employees. Do these distinc¬ tions make sense? 2. Why are financial bonuses based only on the factors cited, despite the comprehensive cate¬ gories on which employees are rated during performance appraisals? 3. How might a team approach be introduced
306 and developed to increase employee involve¬ ment, empowerment, and commitment? What changes might be required in the roles of managers and line workers if the plant was reorganized into teams? 4. What advice might you give the company to
Part 2
Quality in High-Performance Organizations
better leverage its performance appraisal approach? 5. Can you advance some reasons why turnover is increasing and employee dissatis¬ faction growing, based on HR processes and leadership factors?
III. TVS Partnership Proprietary, Ltd., Brisbane, Australia80 81
The environment of property development and the need for architectural services has been very volatile in Australia. In the mid-1990s the develop¬ ment of commercial properties, such as resorts, office buildings, hotels, and apartment complexes, went through a rapid boom and bust cycle. Many developers eventually declared bankruptcy, and as a result, numerous architectural and property development firms either shrank in size or went bankrupt also. TVS Partnership Proprietary, Ltd.,82 is a small, closely held professional partnership that in 1990 had a staff of 12, and grew to about 22 people in the parent organization, plus approximately 55 others in subsidiary firms. From 1990 until 1997, TVS won numerous design profession awards, as well as the Australian Quality Award in 1993, which it was the first small business in Australia to win. To smooth out business cycles, TVS has diver¬ sified into other services, such as interior design, environmental design, landscape architecture, and hotel property management. It invested in research, developing the first solar house (Solar I) in Australia, and pioneered in efforts to use "envi¬ ronmentally friendly" building design and man¬ agement processes. During the period between 1990 and 1997, the firm grew by more than 50 per¬ cent in number of employees. Annual turnover (revenue) increased by approximately 50 percent per year for the last several years. TVS was first introduced to formal concepts of TQ in 1989. The organization is led by five direc¬ tors. All five are involved in, and responsible for, promoting quality, motivation, improvement, plan review, competitive performance, goals and objec¬ tives, education and training, and customer and supplier relations. Individual directors are respon¬ sible for overview and improvement of specific areas of activity, such as finance and administra¬ tion, office facilities, operations, research and development, human resources, public relations.
and quality processes. They are very proud of their leadership style. The vision of TVS focuses on three areas—self, customers, and community. "Success through ser¬ vice" was developed as a company slogan in 1990. The firm broadened its perspective to a customer focus about 1992. In 1993 it added a community focus. The I-CARE philosophy, an acronym for "Improvement that is continuous and relentless," extends across the TVS organization. It expresses values, goals, and aspirations and provides a prac¬ tical and tangible guide to work by individuals within the firm. It is also, of course, a marketing tool that is used to ensure that customers are aware of the philosophy and implied quality that they can expect to receive from the firm. The TVS philosophy can be summarized by: • • • • • • °
• • • • •
Improvement that is continuous and relentless The power of teams The importance of its individual members The long-term focus Understanding and satisfying client needs Creative, cost-effective, and efficient solu¬ tions that give more with less The integration of the specialist design disci¬ plines—architecture (the built environment), interior design (the human interface), and landscape design (the natural environment) The value of efficient and appropriate supporting processes and technology Management by measurement, data, and analysis A long-term commitment to the local and global environment The encouragement of honesty, trust, integrity, and responsibility Innovation
Recruiting is based primarily on referral with little advertising. In addition, TVS extends offers to students who have worked at the firm through
Chapter 6
Human Resource Practices
a cooperative student program at Queensland University of Technology, and selects the best interns who graduate from various university hos¬ pitality programs (for its hotel division). During a probation period of four months the company tries to assess and retain new employees who are high achievers and people with good values. Clarity of induction, through use of an induc¬ tion manual, helps to ensure that all divisions and new employees are aligned with company goals. The I-CARE philosophy and principles are stressed for new hires. TVS enjoys very low turnover; average retention is about seven years. TVS's induction process allows the managing director to walk into any division and immedi¬ ately feel comfortable. The values emphasized are honesty and integrity, customer focus, the concept of customer service, service to the community, fire and enthusiasm, commitment to grow, commit¬ ment to change, support of a dynamic theory of the firm, and understanding of the need for con¬ tinuously changing processes. Some weaknesses are apparent in the induc¬ tion process, however, especially as it relates to transmission of the TQ philosophy. Those who were at TVS at the time of receiving the AQA are probably the most enthusiastic about TQ. Even the veteran draftsmen could explain the quality process and its components. However, this famil¬ iarity seems to be lacking in newer employees. The induction manual was written just prior to receipt of the AQA Award and has not been revised since that date. In it, the current corporate philosophy, culture, teamwork, and the reward and recognition process are mentioned only in passing. The nature of the employee mix is now quite varied, compared to 1992. The firm now has technical staff, administrative staff, service staff, and professionals (architects, designers, managers, etc.). They need general knowledge about the firm and its everyday procedures, but also specialized knowledge pertaining to their division. TVS experienced a serious drop in morale in 1994 due to three factors: (1) an external environ¬ mental slump, with many property developers going bankrupt, (2) a shift of the firm's offices to temporary quarters for six months while a new office was being built, and (3) new offshore expan¬ sion projects. At that point TVS developed a new strategic planning approach. The Leadership team (five directors, two associates, and the quality
307 manager), reorganized the management of the firm into teams, and dropped centralized job con¬ trols. Job costing was shifted to the teams, and an administrative team became facilitators. Quality control was also made a team responsibility. Prior to setting up the teams, weekly align¬ ment meetings for the whole office had been held to coordinate employee and company efforts. Everyone had to speak or pass at the meetings run without meeting facilitators. As the firm grew, even less input took place, because the less articu¬ late staff became reluctant to speak in a largegroup setting. Eventually, these weekly meetings were discontinued because they were becoming too unproductive. Weekly meetings of smaller teams were used as a substitute for the alignment meetings. How¬ ever, weekly meetings were sometimes problem¬ atic because of travel schedules and urgent tasks of directors and team leaders. Of 12 planned monthly meetings, the directors were able to attend only 8 to 10. Teams sometimes became jealous of each other's territories. Although these team meetings produced cohesion among group members, the level of interteam coordination suf¬ fered. When more people were needed on a team, some were moved to the team needing resources (not always happily). TVS is currently finalizing another team reor¬ ganization to improve interteam coordination. Two people from each team will now meet weekly. Full office meetings will be held monthly. The cost coordination will fall on the shoulders of the pro¬ ject leader. Compensation is an area where the directors of TVS have struggled. Before 1993 they set aside 10 percent of profit for salary increases and bonuses. They would meet and debate how much each person should get at each level. They came up with several elaborate schemes for allocating the amounts to each person systematically. In the end, this arrangement did not work well. After winning the AQA, they went back to the old allocation method. TVS tried to diversify the types of reward and recognition systems. One such program. The Eagle Card, was a motivational system developed by an American management consultant. For out¬ standing performance, employees received Eagle Cards and various rewards associated with them. However, TVS determined that such approaches
308 have a short life cycle, after which they must be replaced by a new scheme. TVS's directors are flexible about the type of rewards and recognition used. They like to reward employees for outstanding performance as close to the event as possible. For example, they have done such things as paid for an employee's wedding, paid the deposit for a person's house, given an employee time off and paid for a trip to Bangladesh for humanitarian purposes, and bought a set of tires for an employee who needed them. They will give time off for a sabbatical, time to sort out family life, and so on. The downside is that employees occasionally feel that they did not receive a reward when they deserved one. Although the directors complied with the AQA criteria, which suggest that they should have a consistent, repeatable process for rewarding people, they struggled with the notion, feeling it was not right for their organization. The final suc¬ cess of many projects cannot be measured for a year or more after commencement. TVS tried an employee shareholding plan in 1990. Many employees were not interested and did not appreciate the value of shares. The plan
IV.
Part 2
Quality in High-Performance Organizations
was reviewed and the company directors eventu¬ ally bought back the employees' shares. Problems arose with the valuation and transfer of shares between employees. Recently they began to issue shares to senior managers. It is now being pro¬ posed that the company will issue and sell shares to senior managers and associates on an invitation basis. Discussion Questions
1. How may changes in the organizational structure affected human resource policies? What were the strengths and weaknesses in the human resources management approaches that TVS adopted? 2. Why has the HR function undergone so many changes in a short period of time? What might you recommend? 3. Creative and operations-oriented parts of an organization often view quality from dif¬ ferent points of view. How could managers from the different divisions be encouraged to share their perspectives and knowledge to benefit the whole organization?
CapStar Health Systems: Human Resource
The complete CapStar case study, a fictitious example of a Baldrige application, can be found on the CD-ROM that accompanies this book. If you have not read CapStar's Organizational Profile yet (see Case III in Chapter 3), please do so first. Examine their response to Category 5 in the con¬ text of the leading practices described in this
chapter (you need not consider the actual Baldrige criteria for this activity). What are their strengths? What are their weaknesses and opportunities for improvement? What specific advice, including useful tools and techniques that might help them, would you suggest?
ENDNOTES 1. Robin Yale Bergstrom, "People, Process, Paint," Production, April 1995,48-51. 2. James L. Heskett, W. Earl Sasser, Jr., and Leonard A. Schlesinger, The Service Profit Chain (New York: The Free Press, 1997), 101. 3. "How Companies Satisfy Employees," available at http://www.Fortmie.com, January 7, 2003. See also 'TOO Best Companies to Work For," Fortune, January 20, 2003. 4. Town Hall discussion at the Quest for Excellence Conference, Washington D.C., March 2000. 5. Richard E. Walton, "From Control to Commit¬ ment in the Workplace," Harvard Business Review 63, no.
2 (March/April 1985), 77-84. © by the President and Fel¬ lows of Harvard College; all rights reserved. 6. Lloyd L. Byars and Leslie W. Rue, Human Resource Management, 6th ed. (New York: Irwin/McGraw-Hill, 2000), 6. 7. Richard Blackburn and Benson Rosen, "Total Quality and Human Resources Management: Lessons Learned from Baldrige Award-Winning Companies," Academy of Management Executive 7, no. 3 (1993), 49-66. 8. Blackburn and Rosen (see note 7). 9. Alfie Kolin, No Contest: The Case Against Competi¬ tion (Boston: Houghton Mifflin, 1986).
Chapter 6
Human Resource Practices
10. Jon R. Katzenback and Douglas K. Smith, "The Discipline of Teams," Harvard Business Review (March/April 1993), 111-120.
309
13. Much of the brief history in this section has been adapted from J. M. Juran, "The QC Circle Phenom¬ enon," Industrial Quality Control, January 1967, 329-336. 14. "Platform Approach at Chrysler," Quality '93: Empowering People with Technology, Fortune Advertise¬ ment, September 20,1993.
30. A more comprehensive review of history and the forerunners of quality circles from the early 1900s can be found in William M. Lindsay, "Quality Circles and Par¬ ticipative Work Improvement: A Cross-Disciplinary His¬ tory," in Dennis F. Ray (ed.), Southern Management Association Proceedings (Mississippi State, MS: Missis¬ sippi State University, 1987), 220-222. 31. Joseph J. Gufreda, Larry A. Maynard, and Lucy N. Lytle, "Employee Involvement in the Quality Process," in The Ernst & Young Quality Improvement Consulting Group, Total Quality!: An Executive's Guide for the 1990s (Homewood, IL: Richard D. Irwin, 1990). 32. From materials provided by Mike Simms, former plant manager.
15. Brock Yates, The Critical Path (Boston: Little, Brown and Co., 1996), 76.
33. "It's My Manager, Stupid," Across the Board, Jan¬ uary 2000, 9.
16. Jeremy Main. Quality Wars (New York: The Free Press, 1994), 62.
34. J. M. Juran, Juran on Leadership for Quality: An Executive Handbook (New York: The Free Press, i989), 264. 35. Phillip A. Smith, William D. Anderson, and Stanley A. Brooking, "Employee Empowerment: A Case Study," Production and Inventory Management 34, no. 3 (1993), 45-50.
11. Jack D. Orsburn, Linda Moran, Ed Mussel white, and John H. Zenger, Self-Directed Work Teams (Homewood, IL: Business One-Irwin, 1990), 8. 12. Brian Dumaine, "The Trouble with Teams," For¬ tune, 5 September, 1994, 86-92.
17. Sidney P. Rubinstein, "QC Circles and U.S. Par¬ ticipative Movements," 1972 ASQC Technical Conference Transactions, Washington, D.C., 391-396. 18. For more about the history and impact of quality circles in the early 1980s in the United States, see William M. Lindsay, Measurement of Quality Circle Effec¬ tiveness: A Survey and Critique, unpublished M.S. thesis. University of Cincinnati, College of Engineering (May 1986), 72,117-120. 19. Helene F. Uhlfelder, "It's All About Improving Performance," Quality Progress, February 2000, 47-52. 20. Harvey A. Robbins and Michael Finley, Why Teams Don't Work: What Went Wrong and How to Make it Right (Princeton, NJ: Peterson's/Pacesetter Books, 1995), 14-15. 21. Samuel C. Certo, Modem Management, 9th ed. (Upper Saddle River, NJ:Prentice Hall, 2003), 389. 22. Peter R. Scholtes et al.. The Team Handbook: Hoiv to Use Teams to Improve Quality (Madison, WI: Joiner Associates, Inc., 1988) 6-10-6-22. 23. Nancy Page Cooper and Pat Noonan. "Do Teams and Six Sigma Go Together?" Quality Progress 36, no. 6 (June 2003), 26-27. 24. Portions adapted from Chapter 4, "Motivation Through the Design of Work," in J. R. Hackman and G. R. Oldham, Work Redesign (Reading, MA: AddisonWesley, 1980). 25. Hackman and Oldham, 25 (see note 24). 26. David A Garvin, Managing Quality (New York: The Free Press, 1988), 202-203. 27. Tom J. Peters, Thriving on Chaos: Handbook for a Management Revolution (New York: Alfred A. Knopf, 1988). 28. Alan Wolf, "Golden Opportunities," Beverage World, February 1991. 29. Robert Slater, Jack Welch and the GE Way (New York: McGraw-Hill, 1999), 153-155,158-159.
36. John Troyer, "Empowerment," Guest Editorial, Quality Digest, October 1996, 64. 37. AT&T Quality Steering Committee, Great Perfor¬ mances (AT&T Bell Laboratories, 1991), 39; and William Smitley and David Scott, "Empowerment: Unlocking the Potential of Your Work Force," Quality Digest 14, no. 8 (August 1994), 40-46. 38. "Changing a Culture: DuPont Tries to Make Sure That Its Research Wizardry Serves the Bottom Line," The Wall Street Journal, March 27,1992, A5. 39. Robert S. Kaplan, "Texas Eastman Company," Harvard Business School Case, No. 9-190-039. 40. John F. Akers, "World-Class Quality: Nothing Else Will Do," Quality Progress 24, no. 10 (October 1991), 26-27. 41. Timothy Aeppel, "Not All Workers Find Idea of Empowerment As Neat As It Sounds," The Wall Street Journal, September 8,1997, Al. 42. Ronald Henkoff, "Make Your Office More Pro¬ ductive," Fortune, February 25,1991, 76. 43. "Small Company's Training Policy Yields Big Results," The Human Element (a publication of the Human Development and Leadership Division of the American Society for Quality), 20, no. 1 (Spring 2003). 44. Alan Wolf, "Coors' Customer Focus," Beverage World, March 1991. 45. Bill Wilson, "Quality Training at FedEx," Quality Digest 15, no. 1 (January 1995), 40-43. 46. "Honda of America Launches Computerized Quality Assurance Training," Quality Progress 30, no. 10 (October 1997), 19-20. 47. "Bonus Pay: Buzzword or Bonanza?" Business Week, November 14,1994, 62-64.
310 48. Woodrumm Imberman, "Pay for Performance Boosts Quality Output," HE Solutions, October 1996, 34-36. 49. Quoted from "Statement of Pam Farr, President & COO, The Cabot Advisory Group on behalf of Cabot Advisory Group, Lie on The Rewarding Performance in Compensation Act before the House Committee on Edu¬ cation and the Workforce Subcommittee on Workforce Protections—April 13,1999." 50. Nancy J. Perry, "Here Come Richer, Riskier Pay Plans," Fortune, December 19,1988, 50-58; "The Nucor Story," available at http://www.nucor.com. 51. Frank C. Barnes, "Nucor (A)," in Robert R. Bell and John M. Burnham, Managing Productivity and Change (Cincinnati, OH: South-Western Publishing Company, 1991), 507. 52. Chris Adams, "Nucor Slashes Its Hot-Rolled Steel Prices by 7%," The Wall Street Journal, September 30,1997, A3. 53. Bruce N. Pfau and Steven E. Gross, Innovative Reward and Recognition Strategies in TQM, The Confer¬ ence Board, Report Number 1051,1993. 54. Brian O'Reilly, "The Mechanic Who Fixed Conti¬ nental," Fortune, December 20,1999,176-186. 55. Dawn Anfuso, "L.L. Bean's TQM Efforts Put People Before Processes," Personnel Journal, July 1994, 73-83. 56. "Bonus Pay: Buzzword or Bonanza?" Business Week, November 14,1994, 62-64. 57. Bob Nelson, "Secrets of Successful Employee Recognition," Quality Digest, August 1996, 26-30. 58. Leigh Ann Klaus, "Motorola Brings Fairy Tales to Life," Quality Progress, June 1997, 25-28. 59. "Domino's Pizza, Inc., "Profiles in Quality (Boston: Allyn and Bacon, 1991), 90-93. 60. Michelle Conlin and Kathy Moore, "Photo Essay—SAS," Business Week, June 19, 2000,192-202. 61. Saul W. Gellerman, Motivation in the Real World (New York: Dutton, 1992). 62. James L. Bowditch and Anthony F. Buono, A Primer on Organizational Behavior, 2d ed. (New York: John Wiley & Sons, 1990), 52. 63. Jeremy Main, Quality Wars (New York: The Free Press, 1994), 130. 64. Khalid A. Aldakhilallah and Diane H. Parente, "Redesigning a Square Peg: Total Quality Management Performance Appraisals," Total Quality Management 13, no. 1 (2002), 39-51. 65. George Eckes, "Practical Alternatives to Perfor¬ mance Appraisals," Quality Progress 27, no. 11 (Novem¬ ber 1994), 57-60. 66. Douglas McGregor, "An Uneasy Look at Perfor¬ mance Appraisal," Harvard Business Review, SeptemberOctober 1972; Herbert H. Meyer, Emanuel Kay, and John R. P. French, Jr., "Split Roles in Performance Appraisal," Harvard Business Review, January-February 1965; Harry Levinson, "Appraisal of What Performance?" Harvard
Part 2
Quality in High-Performance Organizations
Business Review, January-February 1965; A. M. Mohrman, Deming Versus Performance Appraisal: Is There a Resolution? (Los Angeles: Center for Effective Organizations, Univer¬ sity of Southern California, 1989). 67. John F. Milliman and Fred R. McFadden, "Toward Changing Performance Appraisal to Address TQM Concerns: The 360-Degree Feedback Process," Quality Management Journal 4, no. 3 (1997), 44r-64. 68. Milliman and McFadden (see note 67). 69. Dick Grote, "The Secrets of Performance Appraisal: Best Practices from the Masters," Across the Board, May 2000, 14-20. 70. Brian S. Morgan and William A. Schiemann, "Measuring People and Performance: Closing the Gaps," Quality Progress, January 1999, 47-53. 71. See John D. Cook, Susan J. Hepworth , Toby D. Wall, and Peter B. Warr, The Experience of Work (London, Academic Press, 1981); and Dale Henderson and Fess Green, "Measuring Self-Managed Workteams," Journal for Quality and Participation, January-February 1997,52-56. 72. Mark R. Hagen, "Teams Expand into Cyber¬ space," Quality Progress, June 1999, 90-93. 73. Nancy Page Cooper and Pat Noonan, "Do Teams and Six Sigma Go Together?" Quality Progress, June 2003, 25-28. 74. Courtesy of TD Industries. Ben Houston, president. 75. Adapted from the L.L. Bean Web site, http:// www.llbean.com and from Chris Ashton, "HR at the Forefront of Change Management at L.L. Bean," Inter¬ national Journal of Retail Distribution Management 26, no. 4-5, April 14,1998,192. Copyright 1998, MCB Univer¬ sity Press, Ltd (UK). Used with permission of L.L. Bean, Inc. and Chris Ashton. 76. Peter F. Drucker. Management Challenges for the 21st Century. (New York: HarperBusiness, 1999), 139. 77. Michelle Conlin, "The Software Says You're Just Average," Business Week, February 25, 2002,126. 78. Xerox Business Products and Systems, Malcolm Baldrige National Quality Award application, 1989. 79. This case was inspired by a Baldrige assessment project by our former students, Leo Chan, Cara Hast¬ ings, and Eric Vaughn. 80. By William M. Lindsay and Arthur Preston, Senior Research Fellow, Queensland University of Tech¬ nology. See William M. Lindsay and Arthur Preston. "Maintaining Quality Through Evolving Strategy: The TVS Partnership" Industrial Management and Data Sys¬ tems, 100, no. 4, (2000), 164—171, for further details. 81. Appreciation is expressed to the TVS Partnership Proprietary, Ltd., and especially to directors Laurie Truce and Mark Thomson, as well as Penny Pinkham, Quality Manager and Administrative Team Leader, for their hos¬ pitality and cooperation in preparation of this case. 82. Application of The TVS Partnership—Architects to the Australian Quality Award Foundation, Australian Quality Award, Small Enterprise, 1993.
Chapter 6
Human Resource Practices
311
BIBLIOGRAPHY Andersen, Bjorn, and Tom Fagerhaug. Performance Management Explained: Designing and Implementing Your State-of-the-Art System. Milwuakee, WI: American Society for Quality, 2002. AT&T Quality Steering Committee. Batting 1000: Using Baldrige Feedback to Improve Your Business. AT&T Bell Laboratories (1992). Great Performances! AT&T Bell Laboratories, 1991. Badracco, Joseph L. Leading Quietly. Boston: Har¬ vard Business School Press, 2002. Bens, Ingrid. Facilitation at a Glance! Cincinnati: AQP, 1999. Blackburn, Richard, and Benjamin Rosen. "Total Quality and Human Resources Management: Lessons Learned from Baldrige Award-Winning Companies." Academy of Management Executive 7, no. 3 (1993), 49-66. Buckingham, Marcus, and Curt Coffman. First, Break All the Rules: What the World’s Greatest Managers Do Dif¬ ferently. New York: Simon and Schuster, 1999. Byrne, John. Chainsaiv: The Notorious Career of Al Dunlop in the Age of Profit-at-Any-Price. New York:
HarperBusiness, 2002. Christison, William L. "Financial Information Is Key to Empowerment." Quality Progress 27, no. 7 (July 1994), 47-48. Eure, Rob. "E-Commerce (A Special Report): The Classroom—On the Job; Corporate E-Learning Makes Training Available Anytime, Anywhere," The Wall Street Journal, March 12, 2001, R33. Galford, Robert, Laurie Broedling, Edward G. Lawler, III, Tim Riley, et al. "Why Doesn't This HR Department Get Any Respect?" Harvard Business Review, 76, no. 2, March/April 1998, 24-40. Hackman, J. Richard. Leading Teams: Setting the Stage for Great Performance. Boston: Harvard Business School Press, 2002. Herzberg, Frederick. Work and the Nature of Man. Cleveland, OH: World, 1966. _."One More Time: How Do You Motivate Employees?" Harvard Business Review 46, (January/ February 1968), 53-62. Kanfer, Ruth. "Motivation Theory in Industrial and Organizational Psychology." In Marvin D. Dunnette and Leaeta M. Hough (eds.). Handbook of Industrial and Orga¬ nizational Psychology, 2nd ed., vol. 1. Palo Alto, CA: Con¬ sulting Psychologists Press, Inc., 1990, 75-170. Katzenbach, Jon R. and Douglas K. Smith. The Wisdom of Teams. New York: HarperBusiness, 2003. Kern, Jill P., John J. Riley, and Louis N. Jones (eds.). Human Resources Management. Quality and Reliability Series, sponsored by the ASQC Human Resources Divi¬ sion. New York: Marcel Dekker, Inc., and Milwaukee: ASQC Quality Press, 1987.
Lewin, Kurt. A Dynamic Theory of Personality. New York: McGraw-Hill, 1935. Lindsay, William M., and Joseph A. Petrick. Total Quality and Organization Development. Boca Raton, FL: CRC/St. Lucie Press, 1997. Locke, E. A., and G. P. Latham. Goal Setting: A Motivational Technique that Works! Englewood Cliffs, NJ: Prentice Hall, 1984. Mayo, Elton. The Human Problems of Industrial Civi¬ lization. Cambridge, MA: Harvard Graduate School of Business, 1946. Messmer, Max. "Rightsizing, Not Downsizing: How to Maintain Quality Through Strategic Staffing." Industry Week, August 3,1993, 23-26. Miner, John B. Theories of Organizational Behavior. Hinsdale, IL: Dryden Press, 1980. Morgan, Ronald B., and Jacke E. Smith. Staffing the New Workplace: Selecting and Promoting for Quality Improvement. Milwaukee, WI: ASQ Press, 1996.
Mohrman, Susan Albers, Ramkrishnan V. Tenkasi, Edward E. Lawler, III, and Gerald E. Ledford, Jr. "Total Quality Management: Practice and Outcomes in the Largest U.S. Firms, " Employee Relations 17, no. 3 (1995), 26-41. Moorhead, Gregory, and Ricky W. Griffin. Organiza¬ tional Behavior: Managing People and Organizations, 6th ed. New York: Houghton-Mifflin Co., 2001. Olian, Judy D., and Sara L. Rynes. "Making Total Quality Work: Aligning Organizational Processes, Performance Measures, and Stakeholders." Human Resource Management, Fall 1991, 303-333. Palmer, Brian, and Mike Ziemlanski "Tapping Into People," Quality Progress, April 2000, 74-79. Petrick, Joseph A., and Diana Furr. Total Quality in Managing Human Resources. Boca Raton, FL: CRC/St. Lucie Press, 1995. Powell, Cash, Jr. "Empowerment, the Stake in the Ground for ABS." Target, January/February 1992. Rubinstein, Sidney P. "Quality and Democracy in the Workplace." Quality Progress 21, no. 4 (April 1988), 25-28. Semerad, James M. "Create a New Learning Environment." APICS—The Performance Advantage, April 1993, 34-37. Scholtes, P. R. The Team Handbook, 2nd ed. Madison, WI: Joiner Associates, 1996. Simmons, David E., Mark A. Shadur, and Arthur P. Preston. "Integrating TQM and HRM, " Employee Rela¬ tions 17, no. 3 (1995), 75-86. Snape, Ed, Adrian Wilkinson, Mick Marchington, and Ted Redman. "Managing Human Resources for TQM: Possibilites and Pitfalls, " Employee Relations 17, no. 3(1995), 42-51. Snell, Scott A., and James W. Dean. "Integrated
312 Manufacturing and Human Resource Management: A Human Capital Perspective." Academy of Management Journal 35, no. 3 (1992), 467-504. Taylor, Frederick W. The Principles of Scientific Man¬ agement. New York: Harper & Row, 1911. Steers, Richard M. Lyman W. Porter, and Gregory A. Bigley. Motivation and Leadership at Work, 6th ed. New York: McGraw-Hill, 1996.
Part 2
Quality in High-Performance Organizations
Tichy, Noel, and Eli Cohen. "The Teaching Organi¬ zation," Training and Development, July, 1998. Walton, Richard E. "From Control to Commitment in the Workplace." Harvard Business Review 63, no. 2 (March/April 1985), 77-85. Yee, William, and Ed Musselwhite. "Living TQM With Workforce 2000." 1993 ASQC Quality Congress Transactions. Boston, 141-146.
/ Process Management QUALITY PROFILES: STMicroelectronics, Inc.-Region Americas, and Boeing Aircraft and Tanker Programs The Scope of Process Management
Leading Practices Product Design Processes
Cost, Manufacturability, and Quality Design Quality and Social Responsibility Streamlining the Product Development Process Designing Processes for Quality
Special Considerations in Service Process Design Projects as Value-Creation Processes
Project Life Cycle Management Process Control
Process Control in Services Process Improvement
Kaizen
Flexibility and Cycle Time Reduction Breakthrough Improvement Process Management in the Baldrige Criteria,
ISO 9000,
and
Six
Sigma
Quality IN Practice: Gold Star Chili: Process Management QUALITY IN Practice: Bringing Process Management to Education Review Questions Discussion Questions Projects, Etc.
CASES
The State University Experience The PIVOT Initiative at Midwest Bank, Part I Stuart Injection Molding Company CapStar Health Systems: Process Management
The "New Economy," as many call it, is revolutionizing business. Despite the dot¬ com crash at the turn of the century, online shopping has matured, and Amazon.com has garnered high marks in the American Customer Satisfaction Index (see Chapter 4). However, the debut of online retailing was fraught with problems. For example, between Thanksgiving and Christmas, 1999, some 22 million shoppers spent more than $5 billion shopping online.1 Traffic on sites such as Yahoo and Kbkids.com grew by 500 percent. Outpost.com, a computer and electronics retailer, sold $2 million of merchandise in one day. However, it wasn't long before Internet message boards were filled with comments like "I doubt I will ever shop again online for Christmas" and other comments unfit to print here. As Fortune magazine noted, "It takes much more than a logo and a Website to run an e-tailing operation. Online retailers aren't so different from brick and mortar 313
314
Part 2
Quality in High-Performance Organizations
stores. They run out of stock, sell damaged merchandise, and hire rude sales help . . . hordes of companies flooded the market. Trouble is, many of them spent heavily to market and promote their brands but scrimped on infrastructure—the unglamorous side of the business, which focuses on delivering products to customers. The results were often disastrous." Amazon.com, for example, initially tried to have suppliers maintain inventory, but it found that it needed to build traditional distribution cen¬ ters around the country to improve customer service and control over the product.2 A. Blanton Godfrey notes that many organizations are "wired for failure"; that is, their processes are not designed effectively or aligned with each other.3 He cites other examples in addition to the problems that confronted e-retailers. One example is overscheduling at airports. During the 4:15 to 4:30 p.m. time slot, 35 arrivals are sched¬ uled in Atlanta, even though in optimal weather conditions the airport can handle only 25 in 15 minutes; with bad weather, this number drops to 17. Another company celebrated its largest sales contract in history only to discover that all qualified sup¬ pliers for critical materials were at capacity. A third example is the unwillingness of departments to work together. For example, when products fail in the plant or in ser¬ vice, it isn't because designers choose components they know will fail; they often have insufficient information about the problems that result from their choices. These observations point to the importance of designing and managing effective processes—such as product design, order entry, manufacturing, distribution, and customer service—throughout the value chain. Deming and Juran observed that the overwhelming majority of quality problems are associated with processes; few are caused by the workers themselves. Rather, management is responsible to design and continuously improve the processes with which individuals work. Actually, it shares this responsibility with the workforce. The former president of Texas Instruments Process management involves Defense Systems & Electronics Group (now planning and administering the part of Raytheon) had a sign in his office that activities necessary to achieve a high sums up these issues nicely: "Unless you level of performance in key business change the process, why would you expect the processes, and identifying opportu¬ results to change?" nities for improving quality and operational performance, and ulti¬ Process management activities help to pre¬ mately, customer satisfaction. vent defects and errors, eliminate waste and redundancy, and thereby lead to better quality and improved company performance through shorter cycle times, improved flexi¬ bility, and faster customer response. Nearly every leading company views process management as a fundamental business activity (see the Quality Profiles on page 312). AT&T, for example, bases its methodology on the following principles: • Process quality improvement focuses on the end-to-end process. • The mind-set of quality is one of prevention and continuous improvement. • Everyone manages a process at some level and is simultaneously a customer and a supplier. • Customer needs drive process quality improvement. • Corrective action focuses on removing the root cause of the problem rather than on treating its symptoms. • Process simplification reduces opportunities for errors and rework. • Process quality improvement results from a disciplined and structured applica¬ tion of the quality management principles.4
Chapter 7
Process Management
315
Quality Profiles STMicroelectronics, Inc—Region Americas, and Boeing Aircraft and Tanker Programs Headquartered in Carrollton, Texas, STMicro¬ electronics, Inc.—Region Americas (ST), a wholly owned subsidiary of a French firm, ranks among the world's top manufacturers of semiconductor integrated circuits, supplying consumer elec¬ tronics, automotive, medical, telecommunica¬ tions, and computer equipment markets. ST competes against about 20 semiconductor manu¬ facturers with broad product lines as well as hun¬ dreds of smaller rivals that serve niche markets. In this industry, missteps in planning and execu¬ tion quickly translate into competitive disadvan¬ tages. ST aims to distinguish itself through advances in technological innovation, increases in the breadth of its product and service offer¬ ings, and continuous improvement in just-intime delivery, fast prototyping, rapid problem resolution, and other areas responsive to cus¬ tomers' high-priority requirements. In 1998, ST initiated a "gung ho" program to promote teaming and employee empowerment, resulting in the redesign of manufacturing work systems and jobs, all with the aim of encouraging and enabling employees to take control of their work. ST is tightly aligned with its parent corpora¬ tion's quest to become the world leader in envi¬ ronmental compliance. In 1997 and 1998, energy used to manufacture silicon wafers declined by 20 percent. Employee satisfaction levels in 1999 exceed the industry composite in 8 of 10 cate¬ gories, and its supplier management program earned "best in class" rating in an independent evaluation of performance in 19 benchmark areas. ST was a 1999 Baldrige Award winner. Boeing Airlift and Tanker (A&T) Programs designs, develops, and produces the C-17 Globemaster 111 airlifter, which is capable of carrying 170,000 pounds and is used by the U.S. Air Force to transport large, heavy cargo to sites around the world. In 1996, A&T signed a $14.2 billion agree¬
ment to deliver 80 C-17s to the Air Force. A few years later, the Defense Department threatened to cancel the C-17 program because of technical problems, cost overruns, and late deliveries. A&T overhauled its operations to become "processfocused and customer-driven," initiating partner¬ ships with customers, unions, and suppliers, and replacing manager-controlled teams with empowered teams of workers. To help it perform to plan, A&T developed a seven-step approach to defining, managing, stabilizing, and improving processes and established performance measures that are indicators of efficiency and the chief dri¬ vers of customer satisfaction: quality, timeliness, and cycle time. Using this process, one team developed a dry sealant to precoat the 1.4 million fasteners used to assemble a C-17 to replace a wet sealant that was difficult to apply and cost more to dispose of than to buy. The innovation reduced rework, improved airframe quality, reduced structural fatigue, and enabled mechanics to work "faster, cleaner, and better." Between 1995 and its winning a Baldrige Award in 1998, A&T maintained an on-time delivery record of 100 percent. Productivity increased from $200,000 per employee in 1994 to more than $300,000 in 1998. Performance on key quality measures has improved by 50 percent during 1994—1998, cycle time was cut by more than 80 percent, and supplier on-time delivery increased from 75.9 percent to 99.8 percent. The C17's 1997 level of performance was nearly four times better than that of the next best competitor's aircraft, and return on net assets was nearly seven times better than the next best competitor.
Source: Malcolm Baldrige National Quality Award, Profiles of Winners, National Institute of Standards and Technology, Depart¬ ment of Commerce.
316
Part 2
Quality in High-Performance Organizations
This chapter discusses philosophies and approaches for designing and managing important processes in an organization. In Chapter 13, we will discuss specific tools and techniques for process improvement in the context of Six Sigma.
THE SCOPE OF PROCESS MANAGEMENT
As we noted in Chapter 1, essentially all work in an organization is performed by some process. Common business processes include acquiring customer and market knowledge, strategic planning, research and development, purchasing, developing new products or services, fulfilling customer orders, managing information, mea¬ suring and analyzing performance, and training employees, to name just a few. Leading companies identify impor¬ Value-creation processes (sometimes called tant business processes throughout core processes) are those most important to "run¬ the value chain that affect customer ning the business" and maintaining or achieving satisfaction. These processes typi¬ cally fall into two categories: valuea sustainable competitive advantage. They drive creation processes and support the creation of products and services, are critical processes. to customer satisfaction, and have a major impact on the strategic goals of an organization. Value-creation processes typically include design, production/delivery, and other critical business processes. Design processes involve all activities that are performed to incorporate customer requirements, new technology, and past learning into the functional specifications of a product (i.e., a manufactured good or service), and thus define its fitness for use. Production/delivery processes create or deliver the actual product; examples are manufacturing, assembly, dispensing medications, teaching a class, and so on. These processes must be designed to ensure that the product will conform to specifications (the manufacturing definition of quality) and also be pro¬ duced economically and efficiently. Product design greatly influences the efficiency of manufacture as well as the flexibility of service strategies, and therefore must be coordinated with production/delivery processes. The ultimate value of the product, and hence, the perceived quality to the consumer, depend on both these types of processes. Support processes are those that are most important to an organization's value-
creation processes, employees, and daily operations. They provide infrastructure for value-creation processes but generally do not add value directly to the product or ser¬ vice. A process such as order entry that might be thought of as a value creation process for one company (e.g., a direct mail distributor) might be considered as a support process for another (e.g., a custom manufacturer). In general, value creation processes are driven by external customer needs while support processes are driven by internal customer needs. Because value creation processes do add value to products and ser¬ vices, they require a higher level of attention than do support processes. Table 7.1 shows the value creation processes and their requirements defined by Pal's Sudden Service. Their support processes include accounting/finance, human resources, maintenance, management information systems, ordering, and stocking. Other critical support processes that lead to business success and growth might be research and development, technology acquisition, supply chain management and supplier partnering, mergers and acquisitions, project management, or sales and marketing. These processes will differ greatly among organizations, depending on the nature of products and services, customer and market requirements, global focus, and other factors.
Process management consists of three key activities: design, control, and improvement.
Chapter 7
Process Management
317
Table 7.1 Value Creation Processes for Pal's Sudden Service Process
Principal Requirements
Order Taking
Accurate, fast, friendly
Cooking
Proper temperature
Product Assembly
Proper sequence, sanitary, correct ingredients
Cash Collection
Accurate, fast, friendly
Slicing
Cut/size, freshness/color
and amounts, speed, proper temperature, neat
Chili preparation
Proper temperature, quantity, freshness
Ham/chicken preparation
Proper temperature, quantity, freshness
Supply chain management
Price/cost, order accuracy
Property acquisition
Sales potential, adherence to budget
Construction
On time, within budget
Marketing & advertising
Clear message, brand recognition
Source: Courtesy of Pal's Sudden Service.
Designing a process begins with identifying and documenting the process. As we stated in Chapter 1, processes generally cut across traditional organizational func¬ tions, and accurately defining a process may take some investigation and thought. Documenting a process involves describing how it is performed. It will likely include developing a process flowchart and writing standard operating procedures and work instructions. A good process design focuses on the prevention of poor quality by ensuring that goods and services meet both external and internal customer require¬ ments, and that the process is capable of achieving the requisite level of performance. The distinction between control and improvement is illustrated in Figure 7.1. Any process performance measure naturally fluctuates around some average level. Abnormal conditions cause an unusual deviation from this pattern. Removing the causes of such abnormal conditions and maintaining level performance is the essence of control. Improvement, on the other hand, means changing the performance to a new level. To apply the techniques of process management, processes must be (1) repeatable, and (2) measurable. Repeatability means that the process must recur over time. The cycle may be long, as with product development processes or patent applications; or it may be short, as with a manufacturing operation or an order entry process. Mea¬ surement provides the ability to capture important quality and performance indicators to reveal patterns about process performance. Each measurement should aim for a standard or target that is driven by customer requirements. Meeting these two condi¬ tions ensures that sufficient data can be collected to reveal useful information for eval¬ uation and control, as well as learning that leads to improvement and maturity. We may view process management according to the three levels of quality dis¬ cussed in Chapter 1. Major value creation and support processes are generally defined at the organizational level; these activities require attention by senior managers. Each major process consists of many subprocesses that are managed by functional managers
318
Part 2
Quality in High-Performance Organizations
Figure 7.1 Control versus Improvement
or cross-functional teams. Finally, each subprocess consists of many specific work steps performed by individuals at the performer level. Boeing Airlift and Tanker (A&T) Pro¬ grams has developed an "enterprise process model" that views the entire business as eight interconnected process families. These major groupings range from enterprise leadership and new business development to production and post-delivery product support. Each family encompasses up to 10 major processes, which, in turn, are made up of several tiers of supporting subprocesses. A&T manages cross-cutting relation¬ ships as "mega-processes" that extend to suppliers and customers. Individuals or groups, known as process owners, are accountable for process per¬ formance and have the authority to manage and improve their process. Process owners may range from high-level executives who manage cross-functional processes to workers who run a manufacturing cell or an assembly operation on the shop floor. Assigning process owners ensures that someone is responsible to manage the process and optimize its effectiveness. Leading Practices
Process management requires a disciplined effort involving all managers and workers in an organization. Companies that are recognized world leaders in quality and customer satisfaction share some common practices. • They define and document important value creation and support processes, and manage them carefully. Many companies use ISO 9000 as a basis for defining and docu¬ menting key processes. Branch-Smith Printing, for example, created more than 40 process maps as part of the process of converting to ISO 9000. Corning Telecommunications Products Division (TPD) has identified and documented more than 800 processes in all areas of its business, of which 50 are designated as "core business processes" that merit special emphasis in continuous improvement efforts. Each core process is owned and managed by a key busi¬ ness leader. Many organizations also recognize that managing supplier rela¬ tionships (i.e., how performance requirements are communicated and ensured, mutual assistance and training, etc.) represents an important support process.
Chapter 7
Process Management
At DaimlerChrysler, for example, suppliers are involved early in the design process. As a result, DaimlerChrysler often finds out about new materials, parts, and technologies before other automakers. • They translate customer requirements into product and service design requirements early in the design process, taking into account linkages between product design requirements and manufacturing or service process requirements, supplier capabilities, and legal and environmental issues. One of the fundamental questions asked by
SSM Health Care during their process design activities is "What are the customei s expected outcomes from the process?" Reviewing patient/customer feedback data, conducting specialized surveys or focus groups, and including customers on design teams help them answer this question. Leading companies coordinate design and production/delivery processes. AT&T Transmission Sys¬ tems has a new product introduction center that evaluates designs based on manufacturing capabilities, recognizing that good designs both reduce the risk of manufacturing defects and improve productivity. The Bell Laboratories engineeiing research center supports the introduction of new processes by simu¬ lating the manufacturing environment needed to evaluate new technologies. An operational policy developed at Eastman Chemical encourages employees to maximize product value by operating the process at target levels, not just within some specification limits, thus better meeting design performance requirements. Eastman Chemical also reviews designs for safety, reliability, waste minimization, patent position, toxicity information, environmental risks, product disposal, and other customer needs. It also conducts a market analysis of key suppliers' abilities to manage costs, obtain materials, maintain produc¬ tion, and ship reliably. • They ensure that quality is built into products and services and use appropriate engi¬ neering and quantitative tools and approaches during the development process.
Eastman Chemical, for instance, uses laboratory modeling of processes, com¬ puter simulation, designed statistical experiments, and evaluation in cus¬ tomers plants to assess the quality of its products prior to production. Texas Instruments locates its design centers strategically throughout its facilities. These centers offer expertise and systems with extensive capability for electrical and mechanical computer-aided design, system engineering, and manufac¬ turing, and allow the evaluation of parts that have the best quality history, producibility, reliability, and other special engineering requirements. AT&T Universal Card Services used qualitative and quantitative research and testing to verify how accurately it understood customer needs. Before it introduced new products or services, market trials were conducted to determine whether they met customer and business requirements. The company's program man¬ agement process had guidelines for deliverables, which addressed all quality requirements. Each phase of the process fulfilled specific requirements, which had to be completed, reviewed, and approved before the next phase of devel¬ opment began. IBM Rochester uses statistical techniques to study customers' priorities and trade-offs; validates this information with customer councils, sat¬ isfaction surveys, and other forms of feedback; and maintains a Software Partner Laboratory in which customers can certify that requirements are being met and that programs will operate correctly on their systems. • They manage the product development process to enhance cross-functional communica¬ tion, reduce product development time, and ensure trouble-free introduction of products and services. Leading companies use cross-functional teams to coordinate all
phases of product development and reduce development times. Boeing A&T
319
320
Part 2
Quality in High-Performance Organizations
has more than 100 integrated product teams (IPTs) that oversee the design, pro¬ duction, and delivery of the C-17 aircraft's more than 125,000 parts and sup¬ porting services. AT&T established nine expert breakthrough teams—called Achieving Process Excellence Teams—that identify process improvements for developing and deploying products faster in the market. They establish stan¬ dards, procedures, and training for cross-functional communication that pre¬ vents problems from occurring. At The Ritz-Carlton Hotel Company, for instance, the interface of all design, marketing, operations, and legal functions throughout each project allows the company to anticipate requirements and evaluate progress. Customized hotel products and services, such as meetings and banquet events, receive the full attention of local hotel cross-functional teams. These teams involve all internal and external suppliers, verify produc¬ tion and delivery capabilities before each event, critique samples, and assess results. At Globe Metallurgical, a team consisting of employees from customer service, engineering, and quality assurance works together before product development even begins. Afterward, a team of customers and employees from purchasing, engineering, and quality assurance works with the first team to manage the development process. To ensure a trouble-free launch of its prod¬ ucts, Solar Turbines uses advanced computerized design and analytical tools that ensure collaboration and sharing of data between manufacturing and key suppliers. Other tools, such as predictive modeling and rapid prototyping are used to validate function, performance, and manufacturability. • They define performance requirements for suppliers, ensure that requirements are met, and develop partnering relationships with key suppliers and other organizations. At Dana Commercial Credit, strategic suppliers include financial institutions and law firms. Legal requirements are communicated at the early stages of a relation¬ ship; feedback from customers determines whether requirements are being met. Corning TPD classifies its suppliers in a hierarchy: Level 1 suppliers have a direct impact on customer satisfaction; Level 2 suppliers are important, but do not have direct linkage to customer satisfaction; Level 3 suppliers provide commodity-like products. Level 1 suppliers are supported by cross-functional teams and inte¬ grated into development activities. Armstrong conducts site visits and has a 5level scale to help suppliers understand where they stand in meeting the company's expectations. STMicroelectronics developed an annual Supplier Quality & Service Plan, which sets goals for suppliers and specifies how ST will review performance, share data, and carry out other responsibilities in the rela¬ tionship. Long-term partnerships with quality-minded suppliers enabled Texas Nameplate Company to nearly eliminate inspections of incoming materials. These "ship-direct-to-stock" suppliers are required to be defect-free for at least two years and meet all requirements specified on purchase orders. • They control the quality and operational performance of key processes and use system¬ atic methods to identify significant variations in operational performance and output quality, determine root causes, make corrections, and verify results. Leading compa¬
nies establish measures and indicators to track quality and operational perfor¬ mance, and use them as a basis for controlling the processes and consistently meeting specifications and standards. The key measures used by SSM Health Care to monitor their processes are shown in Table 7.2. Daily, weekly, monthly, and quarterly performance assessments provide the opportunity to review and manage these measures and identify ways of preventing potential errors before they affect the patient. The Ritz-Carlton Hotel Company has a policy by which the first person who detects a problem is empowered to break away from
Chapter 7
Process Management
321
Table 7.2 SSM Health Care Process Requirements and Measures Process
Key Requirements
Key Measures
Timeliness
• Time to admit patients to the setting of care • Timeliness in admitting/registration rate on patient satis¬ faction survey questions
Timeliness
• % of histories and physicals charted within 24 hrs or prior to surgery
Admit Admitting/ Registration
Assess Patient Assessment
• Pain assessed at appropriate intervals, per hospital policy Clinical laboratory and radiology services
Accuracy & Timeliness
• Quality control results/Repeat rates • Turnaround time • Response rate on medical staff satisfaction survey
Care Delivery/Treatment Provision of clinical care
Nurse responsiveness Pain management Successful clinical outcomes
• Response rate on patient satisfaction and medical staff survey questions • Wait time for pain medications • % CHF patients received med instructions/weighing • % Ischemic heart patients discharged on proven therapies • Unplanned readmits/Returns to ER or Operating Room • Mortality
Pharmacy/ Medication use
Accuracy
• Use of dangerous abbreviations in medication orders • Med error rate or adverse drug events resulting from med errors
Surgical services/ Anesthesia
Professional skill, competence/ communication
Clear documentation of informed surgical and anesthesia consent • Perioperative mortality • Surgical site infection rates
Case management
Appropriate utilization
• Average length of stay (ALOS) • Payment denials • Unplanned readmits
Discharge from setting of care
Assistance and clear directions
• Discharge instructions documented and provided to patient • Response rate on patient satisfaction survey
Discharge
Source: Courtesy of SSM Health Care.
routine duties, investigate and correct the problem immediately, document the incident, and then return to their routine. Many companies use statistical process control (see Chapter 14) and formal problem-solving processes to iden¬ tify, analyze, and solve quality problems. Granite Rock, for instance, was the first in the construction materials industry to apply statistical process control in the management of production of aggregates, concrete, and asphalt products. • They continuously improve processes to achieve better quality, cycle time, and overall operational performance. Leading companies employ systematic approaches for
Part 2
322
Quality in High-Performance Organizations
analyzing data and identifying improvements. Branch-Smith Printing, for example, uses a simple quality improvement process (QIP) shown in Figure 7.2 to evaluate and improve all production and delivery processes, using performance data and complaints to prioritize opportunities for process improvement. Leading companies use proven techniques such as process analysis and simplifi¬ cation and advanced technologies. Motorola's Commercial, Government, and Industrial Solutions Sector uses continuous improvement teams that meet regu¬ larly to proactively evaluate and improve processes. At IBM Rochester, cross¬ functional teams examine all elements of production and support processes, from order entry to delivery and installation. The teams evaluate and remove, change, and improve steps in these processes. The Ritz-Carlton has eight mechanisms devoted solely to the improvement of process, product, and service quality: 1. New hotel start-up improvement process: a cross-sectional team from the entire company that works together to identify and correct problem areas. 2. Comprehensive performance evaluation process: the work area team mecha¬ nism that empowers people who perform a job to develop the job procedures and performance standards. 3. Quality network: a mechanism of peer approval through which an indi¬ vidual employee can advance a good idea. 4. Standing problem-solving team: a standing work area team that addresses any problem it chooses. 5. Quality improvement team: special teams assembled to solve an assigned problem identified by an individual employee or leaders. 6. Strategic quality planning: annual work area teams that identify their mis¬ sions, primary supplier objectives and action plans, internal objectives and action plans, and progress reviews.
Figure 7.2 QIP Process at Branch-Smith Printing
Used with permission of AIM, Inc.
Chapter 7
Process Management
7. Streamlining process: the annual hotel evaluation of processes, products, or services that are no longer valuable to the customer. 8. Process improvement: the team mechanism for corporate leaders, managers, and employees to improve the most critical processes. • Thei/ innovate to achieve breakthrough performance using such approaches as bench¬ marking and reengineering. Briefly, benchmarking is the search for best practices in any company, in any industry, anywhere in the world, and reengineering is the radical redesign of business processes to achieve significant improvements in performance. As an example of benchmarking, when Granite Rock could not find any company that was measuring on-time delivery of concrete, it talked with Domino s Pizza, a worldwide leader in on-time delivery of a rapidly perishable product (a characteristic shared with freshly mixed concrete) to acquire new ideas for measuring and improving its processes. AT&T has a corporate database to share benchmarking information among its business units. The database contains data from more than 100 companies and 250 benchmarking activities for key processes such as hardware and software development, manufacturing, financial planning and budgeting, international billing, and service delivery. AT&T obtains this information from customers, visits to other companies, trade shows and jour¬ nals, professional societies, product brochures, and outside consultants. Pal's Sudden Service uses benchmarking extensively. Managers are continually on the lookout for benchmarking candidates, and each one compiles a running list of potential subjects. Pal s uses benchmarking to obtain meaningful competitive comparisons, new best practices for achieving higher performance goals, or new organizational directions, as well as to constantly remind the entire organization that performance can always be improved. To illustrate the concept of reengineering, Intel Corporation previously used a 91-step process costing thousands of dollars to purchase ballpoint pens—the same process used to purchase forklift trucks! The improved process was reduced to eight steps. In rethinking its purpose as a customer-driven, retail ser¬ vice company rather than a manufacturing company. Taco Bell eliminated the kitchen from its restaurants. Meat and beans are cooked outside the restaurant at central commissaries and reheated. Other food items such as diced tomatoes, onions, and olives are prepared off-site. This innovation saved about 11 million hours of work and $7 million per year over the entire chain.6
PRODUCT DESIGN PROCESSES Companies today face incredible pressures to continually improve the quality of their products while simultaneously reducing costs, to meet ever-increasing legal and envi¬ ronmental requirements, and to shorten product life cycles to meet changing con¬ sumer needs and remain competitive. The ability to achieve these goals depends on a large extent on product design (by which we also imply redesign). The complexity of today's products makes design a difficult activity; a single state-of-the-art integrated circuit may contain millions of transistors and involve hundreds of manufacturing steps. Nevertheless, improved designs not only reduce costs, but increase quality. For example, a network interface card from around 1990 contained about 40 chips; five years later, the entire system board of a Macintosh Performa 5200 had just 19. Fewer components mean fewer points of failure and less chance of assembly error.7 Most companies have some type of structured product development process. Although we tend to equate product development with manufactured goods, it is important to realize that design processes apply to services as well. For example, in
323
Part 2
324
Quality in High-Performance Organizations
the late 1980s, Citibank designed a new mortgage approval procedure that reduced turnaround times from 45 to less than 15 days; FedEx has consistently developed new variations of its package delivery services.8 The typical product development process, shown in Figure 7.3, consists of six phases: 1. Idea Generation. As emphasized in Chapter 4, new or redesigned product ideas should incorporate customer needs and expectations. However, true innova¬ tions often transcend customers' expressed desires, simply because customers may not know what they like until they have it. A good example is Chrysler's decision to develop the minivan, despite research that showed that people balked at such an odd-looking vehicle.9 2. Preliminary Concept Development. In this phase, new ideas are studied for feasi¬ bility, addressing such questions as: Will the product meet customers' require¬ ments? Can it be manufactured economically with high quality? Objective criteria are required for measuring and testing the attributes associated with these questions. One tool for assisting in this and subsequent steps is quality function deployment, which will be described in Chapter 12. 3. Product/Process Development. If an idea survives the concept stage—and many do not—the actual design process begins by evaluating design alternatives and determining engineering specifications for all materials, components, and parts. This phase usually includes prototype testing, in which a model (real or simulated) is constructed to test the product's physical properties or use under actual operating conditions, as well as consumer reactions to the prototypes. For example, in developing the user interface for an automobile navigation system, BMW conducted extensive consumer tests with a keyboard, a rotating push button, and a joystick (the push button was ultimately selected).10 Boeing's 777 jet was built using digital prototypes. Design reviews are frequently con¬ ducted to identify and eliminate possible causes for manufacturing and mar¬ keting problems. In addition to the actual product design, companies develop, test, and standardize the processes used in manufacturing, which include selecting the appropriate technology, tooling, and suppliers, and performing pilot runs to verify results.
Figure 7.3 Structured Product Development Process Idea Generation Preliminary Concept Development Product/Process Development Full-Scale Production Market Introduction Market Evaluation
Chapter 7
Process Management
4. Full-Scale Production. If no serious problems are found, the company releases the product to manufacturing or service delivery teams. 5. Market Introduction. The product is distributed to customers. 6. Mm ket Evaluation. Deming and Juran both advocated an ongoing product development process that relies on market evaluation and customer feedback to initiate continuous improvements. In fact, Deming's introductory lecture to Japanese managers in 1950 contrasted the "old way" of product design—design it, make it, and try to sell it—with a "new way": • Design the product (with appropriate tests). • Make it and test it in the production line and in the laboratory. • Put it on the market. • Test it in service through market research; find out what the user thinks of it, and why the nonuser has not bought it. • Redesign the product, in light of consumer reactions to quality and price.11 This philosophy is one of the key ingredients in a successful TQ culture. Many companies view customers as significant partners in product development, thus integrating market evaluation throughout the process. Ames Rubber Company; for example, uses a four-step approach to product development that maintains close communication with the customer.12 Typically, Ames initiates a new product through a series of meetings with the customer and sales/marketing or the technical services group. From these meetings, management prepares a product brief listing all tech¬ nical, material, and operational requirements. The brief is forwarded to internal departments, such as engineering, quality, and manufacturing. The technical staff then selects materials, processes, and procedures, and submits its selections to the customer. Upon the customer's approval, a prototype is made. Ames delivers the prototype to the customer, who evaluates and tests it and reports results to the com¬ pany. Ames makes the requested modifications and returns the prototype for further testing. This process continues until the customer is completely satisfied. Next, Ames makes a limited preproduction run. Data collected during the run are analyzed and shared with the customer. Upon approval, full-scale production commences. Design approaches often differ depending on the nature of products or services. For example, approaches to designing entirely new products will be unlike those that address minor changes and improvements. Design approaches might consider fac¬ tors such as functional performance, cost, manufacturability, safety, and environ¬ mental impacts. We address some of these issues next.
Cost, Manufacturability, and Quality General Electric found that 75 percent of its manufacturing costs are determined by design. With products in which parts alone represent 65 to 80 percent of the manufac¬ turing cost, design may account for 90 percent or more of the total manufacturing cost. Other companies exhibit similar figures. For Rolls Royce, design determines 80 percent of the final production costs. Simplifying the design can often improve Product design can significantly cost as well as quality. By cutting the number of affect the cost of manufacturing parts, material costs generally go down, inven¬ (direct and indirect labor, materials, and overhead), redesign, warranty, tory levels fall, the number of suppliers shrinks, and field repair; the efficiency by and production time can be shortened. Back in ivhich the product can be manufac¬ the days of dot matrix printers (ask your tured, and the quality of the output. instructor!), IBM, for example, realized many benefits of design simplification. IBM had been
325
326
Part 2
Quality in High-Performance Organizations
buying its dot matrix printers from Seiko Epson Corporation, then the world's lowcost producer. When IBM developed a printer with 65 percent fewer parts that was designed to snap together during final assembly without the use of fasteners, the result was a 90 percent reduction in assembly time and major cost reductions. Many aspects of product design can adversely affect manufacturability and, hence, quality.13 Some parts may be designed with features difficult to fabricate repeatedly or with unnecessarily tight tolerances. Some parts may lack details for self-alignment or features for correct insertion. In other cases, parts so fragile or so susceptible to corrosion or contamination may be damaged in shipping or by internal handling. Sometimes a design simply has more parts than are needed to perform the desired functions, which increases the chance of assembly error. Thus, problems of poor design may show up as errors, poor yield, damage, or functional failure in fab¬ rication, assembly, test, transport, and end use. Designs with numerous parts increase the incidence of part mix-ups, missing parts, and test failures. Parts that are similar but not identical create the possibility that an assembler will use the wrong part. Parts without details to prevent insertion in the wrong orientation lead to more frequent improper assembly. Complicated assembly steps or tricky joining processes can cause incorrect, incomplete, unreliable, or otherwise faulty assemblies. Finally, the designer's failure to consider conditions to which parts will be exposed during assembly such as temperature, humidity, vibration, static electricity, and dust, may result in failures during testing or use. DFM is intended to prevent product Design for manufacturability (DFM) is the designs that simplify assembly oper¬ process of designing a product for efficient pro¬ ations but require more complex and duction at the highest level of quality. expensive components, desigits that simplify component manufacture Table 7.3 summarizes important design zohile complicating the assembly guidelines for improving manufacturability and process, and designs that are simple thus improving quality and reducing costs. Many and inexpensive to produce but industries have developed more specific guide¬ difficult or expensive to service or lines. For example, guidelines for designing support. printed circuit boards include: • Placing all components on the topside of the board • Grouping similar components whenever possible • Maintaining a 0.60-inch clearance for insertable components Design Quality and Social Responsibility
Safety in consumer products represents a major issue in design, and certainly an important part of a company's public responsibilities. Liability concerns cause many companies to forgo certain product development activities. For example. Unison Industries, Inc., of Rockford, Illinois, developed a new solid-state electronic ignition system for piston-engine aircraft. The company dropped the product after prototype testing. Unison says it was sued over crashes involving aircraft on which its products were not even installed. Getting removed from the lawsuits proved costly in itself.14 In a survey of more than 500 chief executives, more than one-third worked for firms that canceled the introduction of products because of liability concerns. Many companies closed plants All parties responsible for design, and laid off workers, and more than 20 percent of manufacture, sales, and service of a the executives said they believed their compa¬ defective product are now liable for nies lost market share to foreign competitors damages. because of product liability costs.
Chapter 7
Process Management
327
Table 7.3 Design Guidelines for Quality Assurance Minimize Number of Parts •
Fewer parts and assembly drawings
• Less complicated assemblies • Fewer parts to hold to required quality characteristics • Fewer parts to fail Minimize Number of Part Numbers • Fewer variations of like parts Design for Robustness (Taguchi method) • Low sensitivity to component variability Eliminate Adjustments • No assembly adjustment errors • Eliminates adjustable components with high failure rates Make Assembly Easy and Foolproof • Parts cannot be assembled wrong • Obvious when parts are missing • Assembly tooling designed into part • Parts are self-securing • No "force fitting" of parts Use Repeatable, Well-Understood Processes • Part quality easy to control • Assembly quality easy to control Choose Parts That Can Survive Process Operations • Less damage to parts • Less degradation of parts Design for Efficient and Adequate Testing • Less mistaking "good" for "bad" product and vice versa
-> Lower volume of drawings and instructions to control -> Lower assembly error rate -> Higher consistency of part quality -> Higher reliability -> Lower assembly error rate -> Higher first-pass yield; less degradation of performance with time -> Higher first-pass yield -> Lower failure rate
-> -> -> ->
Lower Lower Lower Lower
assembly assembly assembly assembly
error error error error
rate rate rate rate
-> Less damage to parts; better serviceability -> Higher part yield -> Higher assembly yield -> Higher yield -> Higher reliability -> Truer assessment of quality; less unnecessary rework
Lay Out Parts for Reliable Process Completion •
Less damage to parts during handling and assembly
Eliminate Engineering Changes on Released Products • Fewer errors due to changeovers and multiple revisions/versions
-> Higher yield; higher reliability
-> Lower assembly error rate
Source: D. Daetz, "The Effect of Product Design on Product Quality and Product Cost," Quality Progress 20, no. 6 (June 1987), pp. 63-67.
According to the theory of strict liability, anyone who sells a product that is defec¬ tive or unreasonably dangerous is subject to liability for any physical harm caused to the user, the consumer, or the property of either.15 This law applies when the seller is in the business of selling the product, and the product reaches the consumer without a substantial change in condition even if the seller exercised all possible care in the preparation and sale of the product. The principal issue is whether a defect, direct or indirect, exists. If the existence of a defect can be established, the manufacturer usu¬ ally will be held liable. A plaintiff need prove only that (1) the product was defective, (2) the defect was present when the product changed ownership, and (3) the defect
328
Part 2
Quality in High-Performance Organizations
resulted in injury. In 1997 Chrysler was ordered to pay $262.5 million in a case involving defective latches on minivans; thus, the economic consequences can be significant. , Attention to design quality can greatly reduce the possibility of product liability claims as well as provide supporting evidence in defense arguments. Liability makes documentation of quality assurance procedures a necessity. A firm should record all evidence that shows the designer established test and monitoring procedures of crit¬ ical product characteristics. Feedback on test and inspection results along with cor¬ rective actions taken must also be documented. Even adequate packaging and handling procedures are not immune to examination in liability suits, because pack¬ aging is still within the manufacturer's span of control. Managers should address the following questions:16 • • • • • •
Is the product reasonably safe for the end user? What could possibly go wrong with it? Are any needed safety devices absent? What kind of warning labels or instructions should be included? What would attorneys call "reasonable foreseeable use"? What are some extreme climatic or environmental conditions for which the product should be tested? • What similarities does the product have with others that may have encountered previous problems? In addition to legal issues, environmental concerns have an unprecedented impact on product and process designs. Hundreds of millions of home and office appliances are disposed of each year. The problem of what to do with obsolete computers is a growing design and technological waste problem today.17 A monitor contains eight pounds of lead; a CPU has another three to five pounds, as well as other hazardous metals, such as mercury. According to a 1997 Carnegie Mellon University study, 150 million dead but not decaying PCs will be buried in U.S. landfills by 2005. In Europe, the European Commission proposed a ban on materials such as lead-based solder in PCs and the imposition of recycling responsibilities on manufacturers beginning in January 2004. Pressures from environmental groups clamoring for "socially responsive" DfE is the explicit consideration of designs, states and municipalities that are run¬ environmental concerns during the ning out of space for landfills, and consumers design of products and processes, who want the most for their money all cause and includes such practices as designing for recyclability and dis¬ designers and managers to look carefully at the assembly. concept of design-for-environment, or DfE.18 DfE offers the potential to create more desir¬ able products at lower costs by reducing disposal and regulatory costs, increasing the end-of-life value of products, reducing material use, and minimizing liabilities. Recy¬ clable products are designed to be taken apart and their components repaired, refur¬ bished, melted down, or otherwise salvaged for reuse. Recyclability appeals to environmentalists as well as city and state officials, each of whom are fighting the effects of waste disposal. At the same time, however, it creates new issues for designers and consumers. For example, designers must strive to use fewer types of materials, such as plastics, with properties that allow for reuse. Business Week cites several U.S. firms already working on or marketing such products, including Whirlpool, 3M, and General Electric.19 The latter's plastics division, which serves the durable goods market, uses only thermoplastics in its products. Unlike many other varieties of plashes, ther¬ moplastics can be melted down and recast into other shapes and products, thus making
Chapter 7
Process Management
them recyclable. Designers must also refrain from using certain methods of fastening, such as glues and screws, in favor of quick connect-disconnect bolts or other such fas¬ teners. These changes in design will have an impact on tolerances, durability, and quality of products. Such design changes affect consumers who will be asked to recycle products (perhaps to recover a deposit), in spite of inconveniences such as transporting them to a recycling center. Repairable products are not a new idea, but the concept lost favor when, in the 1960s and 1970s, the United States became known as the “throwaway society.” Many products are discarded simply because the cost of maintenance or repair is too high when compared with the cost of a new item. Now design for disassembly promises to bring back easy, affordable product repair. For example. Whirlpool Corporation is developing a new appliance designed for repairability, with its parts sorted for easy coding. Thus, repairability has the potential of pleasing customers, who would prefer to repair a product rather than discard it. At the same time, companies are challenged to consider fresh approaches to design that build both cost-effectiveness and quality into the product. For instance, even though it is more efficient to assemble an item using rivets instead of screws, this approach is contrary to a design-for-disassembly philosophy. An alternative might be an entirely new design that eliminates the need for fasteners in the first place. Streamlining the Product Development Process
The importance of speed in product development cannot be overemphasized. To suc¬ ceed in highly competitive markets, companies must churn out new products quickly. In 1990, the former Digital Equipment Corp., for example, was about to launch a new generation of computer disk drives. However, because of product design problems, the product was very late and competitors had already released enhanced technology drives at much lower prices. What could have been a huge win became a great failure.20 Nearly every industry is focused on reducing product development cycles. Whereas automakers once took four to six years to develop new models, most are striving to do it within 24 months. In fact, Toyota's goal is just 18 months! Boeing took 54 months to design its 777 airplane; yet the company would like to reduce it to 10 because the market changes so quickly. The product development process can be improved with various advanced technologies, such as computer-aided design (CAD), computer-aided manufacturing (CAM), flexible manufacturing systems (FMS), and computer-integrated manufacturing (CIM). These technologies automate and link design and manufacturing processes, reducing cycle times as well as removing oppor¬ tunities for human error, thus improving quality. Such automation is a significant factor at Toyota.21 Successful product development demands the involvement and cooperation of many different functional groups within an organization to identify and solve design problems and try to reduce product development and introduction times. All depart¬ ments play crucial roles in the design process. The designer's objective is to design a product that achieves the desired functional requirements. The manufacturing engi¬ neer's objective is to produce it efficiently. The salesperson's goal is to sell the product, and the finance person's goal is to make a profit. Purchasing seeks parts that meet quality requirements. Packaging and disOne of the most significant barriers tribution deliver the product to the customer to efficient product development is in good operating condition. Clearly, all busipoor intraorganizational cooperation. ness functions have a stake in the product; therefore, all should work together.
329
330
Part 2
Quality in High-Performance Organizations
Unfortunately, the product development process often is performed without such cooperation. In many large firms product development is accomplished in a serial fashion, as suggested in Figure 7.3. In the early stages of development, design engi¬ neers dominate the process. Later, the prototype is transferred to manufacturing for production. Finally, marketing and sales personnel are brought into the process. This approach has several disadvantages. First, product development time is long. Second, up to 90 percent of manufacturing costs may be committed before manufac¬ turing engineers have any input to the design. Third, the final product may not be the best one for market conditions at the time of introduction. An approach that alleviates these problems is called concurrent engineering, or simultaneous engineering. Typical benefits include 30 to 70 percent less develop¬ ment time, 65 to 90 percent fewer engineering changes, 20 to 90 percent less time to market, 200 to 600 percent improvement in Concurrent engineering is a process quality, 20 to 110 percent improvement in white in which all major functions collar productivity, and 20 to 120 percent higher involved with bringing a product to return on assets.22 market are continuously involved Concurrent engineering involves multi¬ with product development from functional teams, usually consisting of 4 to 20 conception through sales. Such an members and including every specialty in the approach not only helps achieve company. The functions of such teams are to trouble-free introduction of prod¬ determine the character of the product and ucts and services, but also residts in decide what design methods and production improved quality, lower costs, and methods are appropriate; analyze product func¬ shorter product development cycles. tions so that all design decisions can be made with full knowledge of how the item is supposed to work; perform a design for man¬ ufacturability study to determine whether the design can be improved without affecting performance; formulate an assembly sequence; and design a factory system that fully involves workers. Concurrent engineering has been a major force behind the resurgence of U.S. automobile companies by enabling them to dramatically reduce product develop¬ ment time. In the past, automobile development followed a sequential process in which styling engineers dreamed up a concept and sent the concept to product engi¬ neers to design components. They in turn would send the designs to manufacturing and suppliers. This process was costly and inefficient; each handoff lost something in time and money. What appeared feasible for one group often proved impossible for another to accomplish. By the time the vehicle was finally produced, marketing was faced with selling a product for which they had no input. Often the vehicle was priced incorrectly for the target market. In 1980 Ford launched Team Taurus, modeled after program management con¬ cepts in the aerospace industry. Program managers headed product teams that included representatives from design, engineering, purchasing, marketing, quality assurance, sales, and service. Cadillac adopted simultaneous engineering in 1985. Vehicle teams, composed of disciplines from every area of the organization, were responsible for managing all steps of product development. They defined the target market and the overall vehicle goals, and managed the timing, profitability, and con¬ tinuous improvement of the vehicle's quality, reliability, durability, and performance. Chrysler's adaptation of simultaneous engineering enabled it to develop and intro¬ duce the celebrated Viper sports car in just two years. Among U.S. automakers Chrysler was an innovator in fast product development.23 One approach often used to facilitate product development is the design review. The purpose of a design review is to stimulate discussion, raise questions, and generate
Chapter 7
Process Management
new ideas and solutions to help designers anticipate problems before they occur. Gen¬ erally, a design review is conducted in three major stages: preliminary intermediate, and final. The preliminary design review establishes early communication between marketing, engineering, manufacturing, and purchasing personnel and provides better coordination of their activities. It usually involves higher levels of management and concentrates on strategic issues in design that relate to customer requirements and thus the ultimate quality of the product. A preliminary design review evaluates such issues as the function of the product, conformance to customer's needs, completeness of spec¬ ifications, manufacturing costs, and liability issues. After the design is well established, an intermediate review takes place to study the design in greater detail to identify potential problems and suggest corrective action. Personnel at lower levels of the organization are more heavily involved at this stage. Finally, just before release to production, a final review is held. Materials lists, drawings, and other detailed design information are studied with the purpose of pre¬ venting costly changes after production setup. In summary, a total approach to product development and process design involves the following activities:24 1. Constantly thinking in terms of how one can design or manufacture products better, not just solving or preventing problems 2. Focusing on "things done right" rather than "things gone wrong" 3. Defining customer expectations and going beyond them, not just barely meeting them or just matching the competition 4. Optimizing desirable features or results, not just incorporating them 5. Minimizing the overall cost without compromising quality of function Various tools that fall under the rubric of "Design for Six Sigma" will be discussed in Chapter 12 and contribute to achieving these objectives.
DESIGNING PROCESSES FOR QUALITY
The design of the processes that produce and deliver goods and services can have a significant impact on cost (and hence profitability), flexibility (the ability to produce the light types and amounts of products as customer demand or preferences change), and the quality of the output. Standardized processes establish consistency of output. For example, in producing a new, very small CD player, Sony had to develop entirely new manufacturing processes, because no process in existence was able to make this product as small and as accurate as the design required. FedEx developed a wireless data collection system that employs laser scanners to manage millions of packages daily through its six main hubs, improving not only customer service, but saving labor costs as well.25 However, standardized processes may not be able to meet the needs of different customer segments as we discussed in Chapter 4. Today, many companies use a strategy of mass customization providing personalized, custom-designed products to meet individual customer preferences at prices comparable to mass-produced items. Motorola, for instance, produces one-of-a-kind pagers from more than 29 million com¬ binations of options in a mass assembly process at a low cost. Dell Computer config¬ ures each computer system to customer specifications. At Levi Strauss, customers are measured for custom-fit jeans at local stores; the jeans are produced at a central factory and delivered to the customer's store. Mass customization requires significant changes to traditional manufacturing processes that focus on either customized, crafted products or mass-produced, standardized products.26 These products include
331
Part 2
332
Quality in High-Performance Organizations
flexible manufacturing technologies, just-inThe goal of process design is to time systems, information technology, and an develop an efficient procedure to sat¬ emphasis on cycle time reduction. , isfy both internal and external cus¬ The design of a process begins with the tomer requirements. process owner. A process owner might be an individual, a team, a department, or some cross¬ functional group. A basic approach to process design is suggested by Motorola: 1. 2. 3. 4.
Identify the product or service: What work do I do? Identify the customer: Who is the work for? Identify the supplier: What do I need and from whom do I get it? Identify the process: What steps or tasks are performed? What are the inputs and outputs for each step? 5. Mistake-proof the process: How can I eliminate or simplify tasks? What "pokayoke" (i.e., mistake-proofing) devices (see Chapter 13) can I use? 6. Develop measurements and controls, and improvement goals: How do I evaluate the process? How can I improve further?
Steps 1 through 3 address such questions as "What is the purpose of the process?" "How does the process create customer satisfaction?" and "What are the essential inputs and outputs of the process?" Step 4 focuses on the actual process design by defining the specific tasks performed in transforming the inputs to outputs. Step 5 focuses on making the process efficient and capable of delivering high quality. Step 6 ensures that the process will be monitored and controlled to the level of required performance. This monitoring involves gathering in-process measurements and/or customer feedback on a regular basis and using this information to control and improve the process. The actual process design is the specification of how the process works. The first phase is to list in detail the sequence of steps—value-adding activities and specific tasks—involved in producing a product or delivering a service, usually depicted as a flowchart (see Chapter 13 for further discussion). Such a graphical representation provides an excellent communication device for visualizing and understanding the process. Flowcharts can become the basis for job descriptions, employee-training programs, and performance measurement. They help managers to estimate human resources, information systems, equipment, and facilities requirements. As design tools, they enable management to study and analyze processes prior to implementa¬ tion in order to improve quality and operational performance. Figure 7.4 shows The Ritz-Carlton's Three Steps of Service process. The process is highly structured and defines the procedures for anticipating and complying with customer needs. All employees who come in contact with customers are trained to follow this process. Special Considerations in Service Process Design
The fundamental differences between manufacturing and service processes deserve special attention in process design. This aspect is especially important because support processes are basically services. Some common examples of service processes are Service process designers must con¬ preparing an invoice, taking a telephone order, centrate on doing things right the processing a credit card, and checking out of first time, minimizing process com¬ a hotel. First, the outputs of service processes plexities, and making the process are not as well defined, as are manufactured immune to inadvertent human products. For example, even though all banks errors, particularly during cus¬ offer similar tangible goods such as checking, tomer interactions. loans, automatic tellers, and so forth, the real
Chapter 7
Process Management
333
Figure 7.4 The Ritz-Carlton Hotel Company: Three Steps of Service Process
Source: © 1992, The Ritz-Carlton Hotel Company. All rights reserved.
differentiating factor among banks is the service they provide. Second, most service processes involve a greater interaction with the customer, often making it easier to identify needs and expectations. On the other hand, customers often cannot define their needs for service until after they have some point of reference or comparison. Fast-food restaurants, for example, have carefully designed their processes for high degree of accuracy and fast response timer' New hands-free intercom systems, better microphones that reduce ambient kitchen noise, and screens that display a cus¬ tomer's order are all focused on these requirements. Timers at Wendy's count every segment of the order completion process to help managers identify problem areas. Kitchen workers wear headsets to hear orders as they are placed. Even the use of photos on drive-through order boards make it more likely for customers to select these items; less variety means faster order fulfillment. Service processes often involve both internal and external activities, a factor that complicates quality design. In a bank, for example, poor service can result from the way that tellers treat customers and also from poor quality of computers and commu¬ nications equipment beyond the control of the tellers. Internal activities are primarily concerned with efficiency (quality of conformance), while external activities—with direct customer interaction—require attention to effectiveness (quality of design). All too often, workers involved in internal operations do not understand how their per¬ formance affects the customers they do not see. The success of the process depends on everyone—workers involved in internal as Services have three basic components: well as external activities—understanding that physical facilities, processes, and prothey add value to the customer. cedures; employees' behavior; and Designing a service essentially involves employees professional judgment.2H determining an effective balance of these com¬ ponents. The goal is to provide a service whose
334
Part 2
Quality in High-Performance Organizations
elements are internally consistent and directed at meeting the needs of a specific target market segment. Too much or too little emphasis on one component will lead to problems and poor customer perceptions. For example, too much emphasis on procedures might result in timely and efficient service, but might also suggest insen¬ sitivity and apathy toward the customer. Too much emphasis on behavior might pro¬ vide a friendly and personable environment at the expense of slow, inconsistent, or chaotic service. Too much emphasis on professional judgment might lead to good solutions to customer problems but also to slow, inconsistent, or insensitive service. A useful approach to designing effective services is first to recognize that services differ in the degree of customer contact and interaction, the degree of labor intensity, and the degree of customization. For example, a railroad is low in all three dimen¬ sions. On the other hand, an interior design service would be high in all three dimen¬ sions. A fast-food restaurant would be high in customer contact and labor intensity, but low in customization. Services low in all three dimensions of this classification are more similar to man¬ ufacturing organizations. The emphasis on quality should be focused on the physical facilities and procedures; behavior and professional judgment are relatively unim¬ portant. As contact and interaction between the customer and the service system increases, two factors must be taken into account. In services low in labor intensity, the customer's impression of physical facilities, processes, and procedures is impor¬ tant. Service organizations must exercise special care in choosing and maintaining reliable and easy-to-use equipment. With higher levels of contact and interaction, appropriate staff behavior becomes increasingly important. As labor intensity increases, variations between individuals become more impor¬ tant; however, the elements of personal behavior and professional judgment will remain relatively unimportant as long as the degrees of customization and contact and interaction remain low. As customization increases, professional judgment becomes a bigger factor in the customer's perception of service quality. In services that are high in all three dimensions, facilities, behavior, and professional judgment must be equally balanced. In services, quality standards take the place of the dimensions and tolerances applicable in manufacturing. Examples of standards set by one of the airline industry leaders, Swissair, include: • Ninety percent of calls are answered within 30 seconds. • Ninety percent of passengers are checked in within three minutes of arrival. • Baggage claim time is only 10 minutes between the first and last customer. However, service standards are inherently more difficult to define and measure than manufacturing specifications. They require extensive research into customer needs and attitudes regarding timeliness, consistency, accuracy, and other service require¬ ments, as discussed in previous chapters. Even though many product specifications developed for manufactured products are focused on meeting a target, such as a product dimension, service targets typically are "smaller is better." Thus, the true ser¬ vice standard is zero defects, and any other standards (such as those of Swissair) should be construed as interim standards and targets only. In designing high-quality service processes, consider the following questions:29 • What service standards are already in place? • Which of these standards have been clearly communicated to all service per¬ sonnel? • Have these standards been communicated to the public?
Chapter 7
Process Management
• Which standards require refinement? • What is the final result of the service provided? What should it ideally be? • What is the maximum access time that a patron will tolerate without feeling inconvenienced ? • How long should it take to perform the service itself? • What is the maximum time for completion of service before the customer's view of the service is negatively affected? • At what point does service begin, and what indicator signals the completion of the service? • How many different people must the consumer deal with in completing the service? • What components of the service are essential? Desirable? Superfluous? • What components or aspects of service must be controlled in order to deliver a service encounter of equal quality each time one occurs? • Which components can differ from encounter to encounter while still leading to a total service encounter that meets standards? What products that affect its service performance does a service organization obtain from other sources? As you can see, service process design is not a trivial exercise!
PROJECTS AS VALUE-CREATION PROCESSES
Some organizations are project-focused because of the nature of their work. They tend to deliver unique, one-of-a-kind products or services tailored to the specific In many companies, value creation processes take the form of projects— temporary work structures that start up, Pr°duce products or services, and
needs of an individual customer. Examples include performing clinical trials for pharmaceutical companies, market research studies, consulting, and systems installation. Project management involves all activities associated
with planning, scheduling, and controlling projects. Good project management ensures that an organization's resources are used efficiently and effectively. Such manage¬ ment is particularly important for Six Sigma, because projects generally cut across organizational boundaries and require the coordination of many different depart¬ ments and functions. Traditional project management methodologies were developed before the advent of total quality; hence, TQ approaches were not often incorporated. Such approaches as identifying customer requirements, using a customer-supplier model, teamwork principles, cycle time reduction, and in-process measurements can improve the quality of the result. For example, although each project is unique, many projects have similar underlying processes, and attention to these processes can improve the overall quality of the project effort. To illustrate, consider Custom Research Incorporated (CRI), which conducts unique market research studies for many different organizations. A Cycle Time Task Force identified nine common processes for all marketing research studies: identifi¬ cation of client requirements/expectations, questionnaire design, questionnaire pro¬ gramming, sampling, data collection, data tabulation, report and analysis, internal communication, and client communication. A Process Task Force was formed to map and improve each process. For example, CRI developed a "one-entry system" that eliminates the need to enter data into its computer system more than once, and
336
Part 2
Quality in High-Performance Organizations
allows questionnaires to be tested for validity and reliability, eliminating several pro¬ gramming steps and helping to reduce cycle time. An account team is in charge of every research project. Project-related problems anywhere in the process are recog¬ nized and reported by the team. Team members use their problem-solving skills to determine whether the variation is due to common or special causes, analyze the rea¬ sons for the occurrence, and implement changes that will prevent it from occurring. When each project is completed, the account team completes a Project Quality Recap documenting problems and solutions and rating the performance of internal depart¬ ments. Teams refer to the Recaps on file when they have similar projects or subse¬ quent projects from the same client.31 Organizations such as Custom Research use a pure project organizational structure whereby team members are assigned exclusively to projects and report only to a project manager. This approach makes it easier to manage projects, because project teams can be designed for efficiency by including the right mix of skills; however, it can result in inef¬ ficiencies because of duplication of resources across the organization, for example, having a different information technology support person on each project. However, in a typical manufacturing or service firm, projects are not the major value creation process, but they often charter projects to meet infrequent needs, such as a new facility layout or technology rollout. Such projects cut across organizational boundaries, making communication across the organization difficult, and require more careful man¬ agement approaches. Functional managers may be reluctant to provide the resources, and employees assigned to projects might relegate a project to a lower priority than their daily, functional job, making it difficult for the project manager to control the project. A practical solution to this dilemma is a matrix organizational structure, which "loans" people and other resources to projects while still maintaining functional con¬ trol over them. Project managers coordinate the work across the functions to mini¬ mize duplication of resources and facilitate communication across the organization, but coordination requires that resources be negotiated (see Chapter 5). Such organi¬ zational structures are often used in Six Sigma organizations. A typical Six Sigma project team consists of a project manager, technical consul¬ tant, project champion, external customer or process owner, and the core project team. We discussed Six Sigma project teams in Chapter 6. The key leadership role belongs to the project manager, who is generally trained as a Six Sigma green belt (SSGB) or black belt (SSBB). Project managers are often generalists who have diverse backgrounds and experience and lead the project activities, plan and Hack progress of the work, and provide direction to the project team. In addition, they must manage the relationships and communication among the members of the project team. Thus, the project manager's ability to facilitate is usu¬ Successful project managers have ally more important than his or her ability to four key skills: a bias toward task supervise. The project manager must also have completion, technical and adminis¬ sufficient technical expertise to resolve disputes trative credibility, interpersonal among functional specialists. and political sensitivity, and leader¬ ship ability. Project Life Cycle Management
A project typically unfolds in stages, which can be called a life cycle. Taking a quality perspective, Kloppenborg and Petrick32 defined the stages of the typical qualityfocused project management process as the following: 1. Project Quality Initiation: Define directions, priorities, limitations, and constraints. 2. Project Quality Planning: Create a blueprint for the scope of the project and resources needed to accomplish it.
Chapter 7
Process Management
3. Project Quality Assurance: Use appropriate, qualified processes to meet technical project design specifications. 4. Project Quality Control: Use appropriate communication and management tools to ensure that managerial performance, process improvements, and customer satisfaction is tracked. 5. Project Quality Closure: Evaluate customer satisfaction with project deliverables and assess success and failures that provide learning for future projects and referrals from satisfied customers. These phases of the project life cycle will be discussed in more detail later in the chapter. These basic components are applicable to any project management endeavor, but are directly relatable to Six Sigma design and improvement projects. The roles and accountability of each member of the project team in each stage of the project life cycle are summarized in Table 7.4. Project Quality Initiation Projects are implemented to satisfy some need of a cus¬ tomer or process owner; thus, the first step in managing a project is to clearly define the goals of the project, and when and how they must be accomplished. Initiation also includes identifying a project champion, project manager, and other team mem¬ bers. The customer must be a vital participant in all stages of the process, not just at the beginning and the end. Project Quality Planning All project-management decisions involve four factors: tiine, resources, costs, and performance. Project managers need to know how much time a project should take and when specific activities should be started and completed so that deadlines can be established and progress of the project monitored. They must also determine the resources, such as people and equipment available for the project, and how they should be allocated among the various activities. Projects usually have limited budgets and costs generally depend on the resources expended; thus, they must be monitored and controlled. Project managers seek ways to minimize costs without jeopardizing deadlines. Finally, performance, which can be defined as how well the results of the project meet customer requirements, should be a measurable entity. Software packages, such as Microsoft Project®, incorporate various quantita¬ tive analysis tools for scheduling, budgetary analysis, and tracking factors of time, resources, and costs, and should be selected at this stage. The project-planning process involves determining the set of activities that must be performed, who will do them, how long each is estimated to take, and when they should be completed to meet the organization s goals. The project-planning process consists of the following steps:
1. Project definition. Define the project, its objectives, and deliverables. Determine the activities that must be completed and the sequence required to perform them. 2. Resource planning. For each activity, determine the resource needs: personnel, time, money, equipment, materials, and so on. 3. Project scheduling. Specify a time schedule for each activity. 4. Project tracking and control. Establish the proper control methods to be used for tracking progress. Develop alternative plans in anticipation of problems in meeting the planned schedule. When projects are late, it is often because of failure to perform these four tasks ade¬ quately.
337
338
Part 2
Quality in High-Performance Organizations
Table 7.4 Project Life Cycle Accountability Matrix Role/Stage
Project Quality Initiation
Project Quality Planning
Project Quality Assurance
Project Quality Control
Project Quality Closure
Champion
Select project manager; promote Six Sigma use; align and select project; commit to charter
Determine decision-making authority; commit to plan; allocate resources needed for project success
Conduct external customer communications; mentor project manager; clear obstacles as needed
Conduct external customer communications; mentor project manager; approve or reject process improvements; clear obstacles as needed
Sign off on completed project; recognize and reward participants; assess project to improve system
External Customer (or Process Owner)
Identify and prioritize expectations; commit to charter
Contribute process knowledge; identify customer satisfaction standards and trade-off values; commit to plan
Participate in ongoing communications; assist in obtaining approvals for changes in processes
Confirm ongoing satisfaction level; accept deliverables
Verify when usage training and support are completed; assess project to improve system; ensure that new processes are implemented; sign off
Master Black Belt (Technical Consultant)
Assist in strategic project selection; promote Six Sigma vision, tools, and process
Assist in identifying data collection and analysis needs; provide training resources; ensure that processes are statistically sound
Participate in ongoing communications; mentor project manager; facilitate cross-project sharing and learning
Provide expertise in design of process improvements; support project manager (SSBB and/or SSGB)
Assist in development of management presentations; do project signoffs; ensure that project results are publicized; disseminate best practices and lessons learned
Project Manager (SSBB and/ or SSGB)
Select core team; identify risks; empower performance; commit to charter
Identify customer satisfaction standards and trade-off values; plan for short-term training if needed; develop quality and communications plans; commit to plan
Conduct customer/ management communications; select tools; confirm qualified processes used; oversee data gathering and analysis; manage quality audits and planning
Track progress, critical success factors, and costs versus plan; implement mid¬ course corrections; measure customer satisfaction; manage process improvements
Notify champion of project completion; recognize and reward participants; assess project to improve system
Core Team
Determine team operating principles; flowchart project; identify lessons learned; commit to charter
Plan project; contribute special expertise; identify suppliers; qualify the process; identify data to collect; commit to plan
Use qualified processes; gather data, find root causes; conduct quality audits; plan future work
Measure customer satisfaction; test deliverables; correct defects; endorse deliverables
Provide customer support and training; assess project to improve system
Source: Adapted from Timothy J. Kloppenborg and Joseph A. Petrick, Managing Project Quality (Vienna, VA: Management Concepts, 2003), 11.
Chapter 7
Process Management
Project Quality Assurance Project quality assurance can be thought of as "customer relationship management" while the project is in process. It requires communication, interpersonal, and diplomacy skills on the part of the project manager. He or she must manage upward to the project champion and out to the client, while keeping a firm, but participative, hand on the pulse of team members and those who are actu¬ ally doing the "hands-on" project work. Project quality assurance allows the project manager to estimate how successfully the final "deliverable" will perform, not just whether it will be on time and below budgeted cost. Software packages such as Microsoft Project are not designed to track "deliverable" performance measures, although some of the project tracking data for estimated final costs and estimated completion dates may be of interest to customers. As suggested in Table 7.4, perfor¬ mance tracking is often subjective, but can be quantified using communication processes and customer surveys, tracking and controlling changes in the project plan, and performing regular project reviews or audits. Project Quality Control Project quality control involves systematically reviewing the
time, resources, cost, and performance measures as the project is being carried out. Because of the uncertainty of task times, unavoidable delays, or other problems, pro¬ jects rarely, if ever, progress on schedule. Managers must therefore monitor perfor¬ mance of the project and take corrective action when needed. A typical project control system includes the following: • A project plan covering expected scope, schedule, cost, and performance goals or requirements • A continuous monitoring system that measures the current results or status against the project plan through the use of monitoring tools • A reporting system that identifies deviations from the project plan by means of trends and forecasts • Timely actions to take advantage of beneficial trends or to correct deviations Project Quality Closure Project closeout is one of those mundane but vitally impor¬ tant processes that facilitate future improvement in project management perfor¬ mance. It consists of such steps as the following:
• Ensuring that the project has been signed off by those who must do so • Ensuring that all bills have been paid and all financial records have been com¬ pleted • Ensuring that team members have not only been thanked, but provided for, which may involve following up with recommendations for reassignment to new projects or departments • Ensuring that "lessons learned" are examined and documented, often by per¬ forming a final project audit • Ensuring that project successes and best practices are communicated and dis¬ seminated to other parts of the organization
PROCESS CONTROL
An international study by Landor & Associates, an independent design and image firm, showed conclusively that Coca-Cola is the number one brand in the minds of soft-drink consumers around the world, and affirmed that the company is totally com¬ mitted to quality. Coca Cola has stated, "Our commitment to quality is something for which we will never lose our taste."33 However, in early June, 1999, quite a few people
339
340
Part 2
Quality in High-Performance Organizations
in Europe did when almost 100 Belgian children fell ill after drinking Coca-Cola. This incident caused the Belgian Health Ministry to require Coca-Cola to recall millions of cans of product in Belgium and to cease product distribution. Later, France and the Netherlands also halted distribution of Coke products as the contamination scare spread. It was quickly determined that contaminated carbon dioxide had been used during the carbonation process at the Antwerp bottling facility. According to the offi¬ cial statement from Coca-Cola, "Independent laboratory testing showed that the cause of the of the off-taste in the bottled products was carbon dioxide. That carbon dioxide was replaced and all bottles with off-taste have been removed from the market. The issue affects the taste of the soft drinks only.... The second issue involves an external odor on some canned products. In the case of the Belgian distribution system, a substance used in wood treatment has caused an offensive odor on the out¬ side bottom of the can. Independent analysis determined that the product is safe. The Company, in conjunction with its bottling partner in Belgium, is taking all necessary steps to eliminate this offensive odor."” After two weeks, the company was allowed to begin producing and distributing products in the three countries. Then, at the end of June, Coca-Cola Beverages Poland found that 1,500 bottles of its Bonaqua water product contained mold. This discovery resulted in 246,000 glass bottles being with¬ drawn from the market in Poland, and replaced with plastic containers. Although the Coca-Cola Company acted swiftly to resolve the problems and recover its image and reputation, this case demonstrates the importance of process con¬ trol. Control is the activity of ensuring conformance to the requirements and taking corrective action when necessary to correct problems and maintain stable perfor¬ mance. Not recognizing when contamination occurs in a bottling process for instance, signifies a lack of control. Control charts, which will be discussed thoroughly in Chapter 14, are Process control is important for an important tool for controlling processes. two reasons. First, process control Any control system has three components: methods are the basis for effective (1) a standard or goal, (2) a means of measuring daily management of processes. accomplishment, and (3) comparison of actual results Second, long-term improvements cannot be made to a process unless with the standard, along with feedback to form the the process is first brought under basis for corrective action. Goals and standards are control. defined during planning and design processes. They establish what is supposed to be accom¬ plished. These goals and standards are reflected by measurable quality characteristics, such as dimensions of machined parts, numbers of defectives, customer complaints, or waiting times. For example, golf balls must meet five standards to be considered as conforming to the Rules of Golf: minimum size, maximum weight, spherical sym¬ metry, maximum initial velocity, and overall distance.35 Methods for measuring these quality characteristics may be automated or performed manually by the workforce. Golf balls are measured for size by trying to drop them through a metal ring—a con¬ forming ball sticks to the ring while a nonconforming ball falls through; digital scales measure weight to one-thousandth of a gram; and initial velocity is measured in a spe¬ cial machine by finding the time it takes a ball struck at 98 mph to break a ballistic screen at the end of a tube exactly 6.28 feet away. Measurements supply the information concerning what has actually been accom¬ plished. Workers, supervisors, or managers then assess whether the actual results meet the goals and standards. If not, then remedial action must be taken. For example, workers might check the first few parts after a new production setup (called setup verification) to determine whether they conform to specifications. If not, the worker adjusts the setup. Sometimes this process occurs automatically. For instance.
Chapter 7
Process Management
in the production of plastic sheet stock, thickness depends on temperature. Sensors monitor the sheet thickness; if it begins to go out of tolerance, the system can adjust the temperature in order to change the thickness. However, in many industries, data are collected through some type of manual inspection process. Such processes that rely on visual interpretation of product char¬ acteristics or manual reading of gauges and instruments may encounter error rates of from 10 to 50 percent. This high rate occurs for several reasons: • Complexity: The number of defects caught by an inspector decreases with more parts and less orderly arrangement. • Defect rate: When the product defect rate is low, inspectors tend to miss more defects than when the defect rate is higher. • Inspection rate: The inspector's performance degrades rapidly as the inspection rate increases.36 These factors can be mitigated by using automated technology, or at the very least, minimizing the number of quality characteristics that must be inspected, reducing time pressures, using repeated inspections (if the same item is inspected by several people, a higher percentage of total defects will be caught), and improving the design of the workspace to facilitate the inspection task. Short-term corrective action generally should be taken by those who own the process and are responsible for doing the work, such as machine operators, orderfulfillment workers, and so on. Long-term remedial action is the responsibility of management. The responsibility for control can be determined by checking the three components of control systems. A process owner must have the means of knowing what is expected (the standard or goal) through clear instructions and specifications; they must have the means of determining their actual performance, typically through inspection and measurement; and they must have a means of making corrections if they discover a variance between what is expected of them and their actual perfor¬ mance. If any of these criteria is not met, then the process is the responsibility of man¬ agement, not the process owner. Both Juran and Deming made this important distinction. If process owners are held accountable for or expected to act on problems beyond their control, they become frustrated and end up playing games with management. Juran and Deming stated that the majority of quality problems are management-controllable—the result of common cause variation. For the smaller proportion of problems resulting from special causes, process owners must be given the tools to identify them and the authority to take action. This philosophy shifts the burden of assuring quality from inspection departments and "quality control" personnel to workers on the shop floor and in customer-contact positions. For example, DaimlerChrysler manufactures the PT Cruiser at the company's Toluca Assembly Plant in Mexico. To ensure quality, the Toluca plant verifies parts, processes, fit, and finish every step of the way, from stamping and body to paint and final assembly. The control practices include visual management through quality alert systems, which are designed to call immediate attention to abnormal conditions. The system provides visual and audible signals for each station for tooling, production, maintenance, and material flow.37 Clearly, if incoming materials are of poor quality, then the final product will certainly be In manufacturing, control is usually applied to incoming materials, key no better. In a TQ environment, customers processes, and final products and should not have to rely on heavy inspection services. of purchased items. The burden of supplying high-quality product should rest with the
341
342
Part 2
Quality in High-Performance Organizations
suppliers themselves. Occasional inspection might be used to audit compliance, but suppliers should be expected to provide documentation and statistical evidence that they are meeting required specifications. The Bonus Materials folder on the CD-ROM contains supplementary material on supplier and partnering process management. Because unwanted variation can arise during production, in-process control is needed throughout the production process. When the process owner assumes the role of inspector, the occurrence of special causes of variation can quickly be recog¬ nized and immediate adjustments to stabilize the process can be made. Done prop¬ erly, this activity can eliminate the need for independent inspection. Final inspection represents the last point in the manufacturing process at which the producer can verify that the product meets customer requirements, and avoid external failure costs. For many consumer products, final inspection consists of functional testing. For instance, a manufacturer of televisions might do a simple test on every unit to make sure it operates properly. However, the company might not test every aspect of the television, such as picture sharpness or other characteristics. These aspects might already have been
evaluated
through
in-process
controls.
Computerized test equipment is quite wide¬ spread, allowing for 100 percent inspection to be conducted rapidly and cost-effectively. Cincinnati Fiberglass, a small manufacturer of fiberglass parts for trucks, uses a control plan for each production process that includes the process name, tool used, standard operating procedure,
tolerance,
inspection
frequency,
Effective quality control systems include documented procedures for all key processes; a clear under¬ standing of the appropriate equip¬ ment and working environment; methods for monitoring and controlling critical quality characteristics; approval processes for equipment; criteria for workmanship, such as written standards, samples, or illus¬ trations; and maintenance activities.
sample size, person responsible, reporting docu¬ ment, and reaction plan. Of particular importance is the ability to trace all compo¬ nents of a product back to key process equipment and operators and to the original material from which it was made. Process control also includes monitoring the accu¬ racy and variability of equipment, operator knowledge and skills, the accuracy of measurement results and data used, and environmental factors such as time and temperature. An example of process control in the food industry—the HACCP approach—is discussed in the Bonus Materials folder on the CD-ROM. Control should be the foundation for organizational learning. Many companies are adopting an approach that has been used in the U.S. military, called after-action review, or debrief. This review consists of asking four basic questions: 1. What was supposed to happen? 2. What actually happened? 3. Why was there a difference? 4. What can we learn? Thus, rather than simply correcting unacceptable events, the focus is on preventing them from occurring again in the future. Process Control in Services Many people think that process control applies only to manufacturing. This assump¬ tion could not be further from the truth. The approach used by The Ritz-Carlton Hotel Company to control quality is proactive because of their intensive personalized sendee Hotel Company
environment.
Systems for collecting and using quality-related measures are widely
deployed and used extensively throughout the organization. Each hotel tracks service quality indicators on a daily basis. The Ritz-Carlton recognizes that many customer
Chapter 7
Process Management
requirements are sensory, and thus, difficult to measure. However, by selecting, training, arid certifying employees in their knowledge of The Ritz-Carlton Gold Stan¬ dards of service, they are able to assess their work through appropriate sensory mea¬ surements—taste, sight, smell, sound, and touch—and take appropriate actions. The company uses three types of control processes to deliver quality: 1. Self-control of the individual employee based on their spontaneous and learned behavior. 2. Basic control mechanism, which is carried out by every member of the work¬ force. The first person who detects a problem is empowered to break away from routine duties, investigate and correct the problem immediately, document the incident, and then return to their routine. 3. Critical success factor control for critical processes. Process teams use customer and organizational requirement measurements to determine quality, speed, and cost performance. These measurements are compared against benchmarks and customer satisfaction data to determine corrective action and resource allocation. In addition. The Ritz-Carlton conducts both self-audits and outside audits. Self audits are carried out internally at all levels, from one individual or function to an entire hotel. Process walk-throughs occur daily in hotels while senior leaders assess field operations during formal reviews at various intervals. Outside audits are per¬ formed by independent travel and hospitality rating organizations. All audits must be documented, and any findings must be submitted to the senior leader of the unit being audited. They are responsible for action and for assessing the implementation and effectiveness of recommended corrective actions. An example of a structured quality control process in the service industry is the "10-Step Monitoring and Evaluation Process" set forth by the Joint Commission on Accrediting Health Care Organizations. This process, shown in Table 7.5, provides a detailed sequence of activities for monitoring and evaluating the quality of health care in an effort to identify problems and improve care. Standards and goals are defined in steps 2 through 5; measurement is accomplished in step 6; and comparison and feedback is performed in the remaining steps. The most common quality characteristics in services, time (waiting time, service time, delivery time) and number of nonconformances, can be measured rather easily. Insurance companies, for example, measure the time to complete different trans¬ actions such as new issues, claim payments, and cash surrenders. Hospitals measure the percentage of nosocomial infections and the percentage of unplanned re-admissions to the emergency room, intensive care, or operating room within, say, 48 hours. Other quality characteristics are observable. They include the types of errors (wrong kind, wrong quantity, wrong delivery date, etc.) and behavior (courtesy, promptness, compe¬ tency, and so on). Hospitals might monitor the completeness of medical charts and the quality of radiology readings, measured by a double-reading process. Simple data collection procedures capture the measurements for service quality con¬ trol. Time is easily measured by taking two observations: starting time and finishing time. Many observed data assume only "yes" or "no" values. For example, a survey of pharmaceutical operations in a hospital might include the following questions: • Are drug storage and preparation areas within the pharmacy under the super¬ vision of a pharmacist? • Are drugs requiring special storage conditions properly stored? • Are drug emergency boxes inspected on a monthly basis? • Is the drug emergency box record book filled out completely?
343
344
Part 2
Quality in High-Performance Organizations
Table 7.5 10-Step Monitoring and Evaluation Process for Health Care Organizations •
•
•
•
•
•
•
•
•
•
Step 7; Assign Responsibility. The emergency department director is responsible for, and actively participates in, monitoring and evaluation. The director assigns respon¬ sibility for the specific duties related to monitoring and evaluation. Step 2: Delineate Scope of Care. The department considers the scope of care provided within emergency services to establish a basis for identifying important aspects of care to monitor and evaluate. The scope of care is a complete inventory of what the emergency department does. Step 3: Identify Important Aspects of Care. Important aspects of care are those that are high-risk, high-volume, and/or problem-prone. Staff identify important aspects of care so that monitoring and evaluation focuses on emergency department activities with the greatest impact on patient care. Step 4: Identify Indicators. Indicators of quality are identified for each important aspect of care. An indicator is a measurable variable related to a structure, process, or outcome of care. Examples of possible indicators (all of which would need to be fur¬ ther defined) include insufficient staffing for sudden surges in patient volume (struc¬ ture), delays in physicians reporting to the emergency room (process), and transfusion errors (outcome). Step 5: Establish Thresholds for Evaluation. A threshold for evaluation is the level or point at which intensive evaluation of care is triggered. A threshold may be 0% or 100% or any other appropriate level. Emergency department staff should establish a threshold for each indicator. Step 6: Collect and Organize Data. Appropriate emergency department staff should collect data pertaining to the indicators. Data are organized to facilitate comparison with the thresholds for evaluation. Step 7: Evaluate Care. When the cumulative data related to an indicator reach the threshold for evaluation, appropriate emergency department staff evaluate the care provided to determine whether a problem exists. This evaluation, which in many cases will take the form of peer review, should focus on possible trends and perfor¬ mance patterns. The evaluation is designed to identify causes of any problems or methods by which care or performance may be improved. Step 8: Take Actions to Solve Problems. When problems are identified, action plans are developed, approved at appropriate levels, and enacted to solve the problem or take the opportunity to improve care. Step 9: Assess Actions and Document Improvement. The effectiveness of any actions taken is assessed and documented. Further actions necessary to solve a problem are taken and their effectiveness is assessed. Step 10: Communicate Relevant Information to the Organization-wide Quality Assur¬ ance Program. Findings from and conclusions of monitoring and evaluation, including actions taken to solve problems and improve care, are documented and reported monthly through the hospital's established channels of communication.
Source: "Medical Staff Monitoring and Evaluation—Departmental Review," Chicago. Copyright by the Joint Com¬ mission on Accreditation of Health Care Organizations, Oakbrook Terrace, IL. Reprinted with permission (undated).
Even though human behavior is easily observable, the task of describing and clas¬ sifying the observations is far more difficult. The major obstacle is developing opera¬ tional definitions of behavioral characteristics. For example, how does one define courteous versus discourteous, or understanding versus indifference? Defining such distinctions is best done by comparing behavior against understandable standards. For instance, a standard for "courtesy" might be to address the customer as "Mr." or
Chapter 7
Process Management
345
"Ms." Failure to do so is an instance of an error. "Promptness" might be defined as greeting a customer within five seconds of entering the store, or answering letters within two days of receipt. These behaviors can easily be recorded and counted. Figure 7.5 shows some behavioral questions used in a patient survey by a group of Southern California hospitals.39
PROCESS IMPROVEMENT
Process improvement is an important business strategy in competitive markets because • Customer loyalty is driven by delivered value. • Delivered value is created by business processes.
Figure 7.5 Sample Hospital Staff Behavior Questions Admissions 11. Altogether, how long did you have to wait to be admitted? More than 1 hour:_(1) 1 hour:_(2) 30 min.:_(3) 15 min.:_ (4) 12. If you had to wait 30 minutes or longer before someone met with you, were you told why? YES:_(1) NO:_(2) Did not wait 30 minutes: (3) Nursing Staff 21. Did a nurse talk to you about the procedures for the day? Never:_(1) Sometimes:_(2) Often:_(3) 22. Were you on IV fluids? YES: _(1) NO:__(2) A. If YES, did the IV fluids ever run out? YES:_(1) NO:_(2) Medical Staff 28. Did the doctor do what he/she told you he was going to do? Never:_(1) Sometimes:_(2) Often:_(3) Housekeeping 36. Did the housekeeper come into your room at least once a day? YES:_(1) NO:_(2) 39. Was the bathroom adequately supplied? Always:_(1) Often:_(2) Sometimes:__(3)
Always:_(4)
Always:
(4)
Never:_(4)
X-Ray When you received services from the X-ray technician, were the procedures explained to you? Always:__(1) Often:_(2) Sometimes:_(3) Never:_(4) Food 34. Generally, were your meals served at the same time each day? Always:_(1) Often:_(2) Sometimes:_(3)
Never:_(4)
Source: Adapted from K. M. Casarreal, J. L. Mill, and M. A. Plant, "Improving Service Through Patient Surveys in a Multihospital Organiza tion," Hospital & Health Services Administration, Health Administration Press, Ann Arbor, Ml, March/April 1986, 41-52. © 1986, Foundation of the American College of Health Care Executives.
346
Part 2
Quality in High-Performance Organizations
• Sustained success in competitive markets requires a business to continuously improve delivered value. • To continuously improve value creation ability, a business must continuously improve its value creation processes.40
/
Microsoft
Improvement should be a proactive A good illustration is Dell. Although it has had task of management and be viewed some of the highest quality ratings in the PC as an opportunity, not simply as a industry, CEO Michael Dell became obsessed reaction to problems and competi¬ with finding ways to reduce machine failure tive threats. rates. He concluded that failures were related to the number of times a hard drive was handled during assembly, and insisted that the number of "touches" be reduced from an existing level of more than 30 per drive. Production lines were revamped and the number was reduced to fewer than 15. Soon after, the reject rate of hard drives fell by 40 percent and the overall failure rated dropped by 20 percent.41 Another example is Microsoft. In 1996, CEO Bill Gates noticed that Microsoft was printing 350,000 sales reports per year, and that 114 different forms were used in procurement alone.42 After many discussions with other top managers and employees, a directive was issued that basically stated that all paper forms and reports must be eliminated unless a compelling need could be found for them. The results included: • The total number of paper forms at Microsoft was reduced from 1,000 to 60; of these, 10 are required by law, 40 are required by outside parties, and 10 are seldom used. • Only one procurement form now exists. • Savings for the first year (1997-98) were estimated at $40 million. • Studies by accounting firms suggest that the average form costs $145 to process, and that the average cost of a similar average electronic transaction, as verified by Microsoft, is $5. Microsoft accomplished this process improvement by using the company intranet, "Frequently Asked Questions" (FAQ), search capabilities, and links to related pages for each electronic form; scanning outside documents and putting them into the internal system; and developing a self-service approach so that individuals could handle 90 percent of their administrative information processing needs on their own desktop PCs. Many opportunities for improvement exist, including the obvious reductions in manufacturing defects and cycle times. Organizations should also consider improving employee morale, satisfaction, and cooperation; improving managerial practices; improving the design of products with features that better meet customers' needs, and that can achieve higher performance, higher reliability, and other marketdriven dimensions of quality; and improving the efficiency of manufacturing sys¬ tems by reducing workers' idle time and unnecessary motions, and by eliminating unnecessary inventory, unnecessary transportation and material handling, and scrap and rework. The concept of continuous improvement dates back many years. One of the ear¬ liest examples in the United States was at National Cash Register Company (NCR). After a shipment of defective cash registers was returned in 1894, the company's founder discovered unpleasant and unsafe working conditions. He made many changes, including better lighting, new safety devices, ventilation, lounges, and lockers. The company offered extensive evening classes to improve employees' edu¬ cation and skills, and instituted a program for soliciting suggestions from factory
Chapter 7
Process Management
workers. Workers received cash prizes and other recognitions for their best ideas; by the 1940s the company was receiving an average of 3,000 suggestions each year. Over the years, many other companies such as Lincoln Electric and Procter & Gamble developed innovative and effective improvement approaches. However, many of these focused almost exclusively on productivity and cost. A focus on quality improvement, on the other hand, is relatively recent, stimulated by the suc¬ cess of the Japanese. Toshiba in 1946, Matsushita Electric in 1950, and Toyota in 1951 initiated some of the earliest formal continuous improvement programs. Toyota, in particular, pioneered just-in-time (JIT), which showed that companies could make products efficiently with virtually zero defects. JIT established a philosophy of improvement, which the Japanese call kaizen (pronounced kl-zen). Kaizen43 Kaizen, which is a Japanese word that means gradual and orderly continuous improve¬ ment, is a philosophy that encompasses all business activities and everyone in an orga¬ nization. In the kaizen philosophy, improvement in all areas of business—cost, meeting delivery schedules, employee safety and skill development, supplier relations, new product development, or productivity—serve to enhance the quality of the firm. Thus, any activity directed toward improvement falls Kaizen focuses on small, gradual, under the kaizen umbrella. Activities to estab¬ and frequent improvements over the lish traditional quality control systems, install long term with minimum financial robotics and advanced technology, institute investment, and participation by employee suggestion systems, maintain equip¬ everyone in the organization. ment, and implement just-in-time production systems all lead to improvement. At Nissan Motor Co., Ltd., for instance, management seriously considers any sug¬ gestion that saves at least 0.6 seconds in a production process. The concept of kaizen is so deeply ingrained in the minds of both managers and workers that they often do not even realize they are thinking in terms of improvement. The Kaizen Institute (http://www.kaizen-institute.com) suggests some basic tips for implementing kaizen. These suggestions include discarding conventional fixed ideas; thinking of how to do something, not why it cannot be done; not seeking perfection; not making excuses, but questioning current practices; and seeking the "wisdom of ten people rather than the knowledge of one." By instilling kaizen into people and training them in basic quality improvement tools, workers can build this philosophy into their work and continually seek improvement in their jobs. This process-oriented approach to improvement encourages constant communication among workers and managers. Three things are required for a successful kaizen program: operating practices, total involvement, and training.44 First, opera ting practices expose new improvement opportunities. Practices such as just-in-time reveal waste and inefficiency as well as poor quality. Second, in kaizen, every employee strives for improvement. Top man¬ agement, for example, views improvement as an inherent component of corporate strategy and provides support to improvement activities by allocating resources effectively and providing reward structures that are conducive to improvement. Middle management can implement top management's improvement goals by estab¬ lishing, upgrading, and maintaining operating standards that reflect those goals; by improving cooperation between departments; and by making employees conscious of their responsibility for improvement and developing their problem-solving skills through training. Supervisors can direct more of their attention to improvement rather than "supervision," which, in turn, facilitates communication and offers better
347
Part 2
348
Quality in High-Performance Organizations
guidance to workers. Finally, workers can engage in improvement through sugges¬ tion systems and small group activities, self-development programs that teach prac¬ tical problem-solving techniques, and enhanced job performance skills. All these improvements require significant training, both in the philosophy and in tools and techniques. The kaizen philosophy has been widely adopted and is used by many firms in the United States and around the world. For example, at ENBI Corporation, a New York manufacturer of precision metal shafts and roller assemblies for the printer, copier, and fax machine markets, kaizen projects have resulted in a 48 percent increase in productivity, a 30 percent reduction in cycle time, and a 73 percent reduction in inventory.45 Kaizen has been successfully applied in the Mercedes-Benz truck factory in Brazil, resulting in reductions of 30 percent in manufacturing space, 45 percent in inventory, 70 percent in lead time, and 70 percent in setup time over a three-year period. Sixteen employees have full-time responsibility for kaizen activities.46 Flexibility and Cycle Time Reduction
Gamble
Success in globally competitive markets requires a capacity for rapid change and flex¬ ibility. Electronic commerce, for instance, requires more rapid, flexible, and cus¬ tomized responses than traditional market outlets. Flexibility might demand special strategies such as modular designs, sharing components, sharing manufacturing lines, and specialized training for employees. It Flexibility refers to the ability to also involves outsourcing decisions, agreements adapt quickly and effectively to with key suppliers, and innovative partnering changing requirements. It might arrangements. mean rapid changeover from one One important business metric that comple¬ product to another, rapid response ments flexibility is cycle time. Cycle time refers to changing demands, or the ability to the time it takes to accomplish one cycle of a to produce a wide range of cus¬ process (e.g., the time from when a customer tomized services. orders a product to the time that it is delivered, or the time to introduce a new product). Reductions in cycle time serve two purposes. First, they speed up work processes so that customer response is improved. Second, reductions in cycle time can only be accomplished by streamlining and simplifying processes to eliminate non-value-added steps such as rework. This approach forces improvements in quality by reducing the potential for mistakes and errors. By reducing non-value-added steps, costs are reduced as well. Thus, cycle time reduc¬ tions often drive simultaneous improvements in organization, quality, cost, and pro¬ ductivity. Significant reductions in cycle time cannot be achieved simply by focusing on individual subprocesses; cross-functional processes must be examined all across the organization. Through these activities, the company comes to understand work at the organizational level and to engage in cooperative behaviors. One example of cycle time reduction is Procter & Gamble's over-the-counter (OTC) clinical division, which conducts clinical studies that involve testing drugs, health care products, or treatments in humans.47 Such testing follows rigorous design, conduct, analysis, and summary of the data collected. P&G had at least four different ways to perform a clinical study and needed to find the best way to meet its research and development needs. They chose to focus on cycle time reduction. Their approach built on fundamental TQ principles: focusing on the customer, fact-based decisions, continual improvement, empowerment, the right leadership structure, and an under¬ standing of work processes. An example is shown in Figure 7.6. The team found that final reports took months to prepare. Only by mapping the existing process did they
Chapter 7
Process Management
349
Figure 7.6 Final Report "Is" and "Should" Process Map Example How a Final Report Is Actually Prepared
How a Final Report Should Be Prepared
Source: David A. McCamey, Robert W. Bogs, and Linda M. Bayuk, "More, Better, Faster From Total Quality Effort," Quality Progress, August 1999, 43-50. © 1999, American Society for Quality. Reprinted with permission.
Part 2
350
Quality in High-Performance Organizations
fully understand the causes of long production times and the amount of rework and recycling during review and sign-off. By restructuring the activities from sequential to parallel work and identifying critical measurements to monitor the process, they were able to reduce the time to less thart four weeks. Agility is a term that is commonly used to characterize flexibility and short cycle times. Agility is crucial to such customer-focused strategies as mass customization, which requires rapid response and flexibility to changing consumer demand. Enablers of agility include close relationships with customers to understand their emerging needs and requirements, empowering employees as decision makers, effective manufacturing and information technology, close supplier and partner rela¬ tionships, and breakthrough improvement (discussed next). Breakthrough Improvement Breakthrough improvement refers to discontinuous change, as opposed to the
gradual, continuous improvement philosophy of kaizen. Breakthrough improve¬ ments result from innovative and creative thinking; often these are motivated by stretch goals, or breakthrough objectives. Stretch goals force an organization When a goal of 10 percent improvement is to think in a radically different way, set, managers or engineers can usually meet it and to encourage major improve¬ with some minor improvements. However, ments as well as incremental ones. when the goal is 1,000 percent improvement, employees must be creative and think "outside of the box." The seemingly impossible is often achieved, yielding dramatic improve¬ ments and boosting morale. Motorola's Six Sigma thrust was driven by a goal of improving product and services quality ten times within two years, and at least 100fold within four years. For stretch goals to be successful, they must derive unambiguously from corpo¬ rate strategy. Organizations must not set goals that result in unreasonable stress to employees or punish failure. In addition, they must provide appropriate help and tools to accomplish the task. Two approaches for breakthrough improvement that help companies achieve stretch goals are benchmarking and reengineering. Benchmarking The development and realization of improvement objectives, partic¬
ularly stretch objectives, is often aided through the process of benchmarking. Bench¬ marking is defined as "measuring your performance against that of best-in-class companies, determining how the best-in-class achieve those performance levels, and using the information as a basis for your own company's targets, strategies, and implementation,"48 or more simply, "the search of industry best practices that lead to superior performance."41' The term best practices refers to approaches that produce exceptional results, are usually innovative in terms of the use of technology or human resources, and are recognized by customers or industry experts. Through benchmarking, a company discovers its strengths and weaknesses and those of other industry leaders and learns how to incorporate the best practices into its own operations. Benchmarking can provide motivation to achieve stretch goals by helping employees to see what others can accomplish. For example, to meet a stretch target of reducing the time to build new 747 and 767 airplanes at Boeing from 18 months (in 1992) to 8 months, teams studied the world's best producers of everything from computers to ships. By 1996 the time had been reduced to 10 months.50 The concept of benchmarking is not new.51 In the early 1800s Francis Lowell, a New England industrialist, traveled to England to study manufacturing techniques
Chapter 7
Process Management
at the best British mill factories. Henry Ford created the assembly line after taking a tour of a Chicago slaughterhouse and watching carcasses, hung on hooks mounted on a monorail, move from one workstation to another. Toyota’s just-in-time produc¬ tion system was influenced by the replenishment practices of U.S. supermarkets. Modern benchmarking was initiated by Xerox and has since become a common prac¬ tice among leading firms. An organization may decide to engage in benchmarking for several reasons. It eliminates "reinventing the wheel" along with associated wasted time and resources. It helps identify performance gaps between an organization and competitors, leading to realistic goals. It encourages employees to continuously innovate. Finally, because it is a process of continuous learning, benchmarking emphasizes sensitivity to the changing needs of customers.52 Three major types of benchmarking have emerged in business. Competitive benchmarking involves studying products, processes, or business performance of
competitors in the same industry to compare pricing, technical quality, features, and other quality or performance characteristics of products and services. For example, a television cable company might compare its customer satisfaction rating or service response time to other cable companies; a manufacturer of TVs might compare its unit production costs or field failure rates against competitors. Significant gaps sug¬ gest key opportunities for improvement. Competitive benchmarking was refined into a science by Xerox during the 1970s and 1980s. Process benchmarking emerged soon after. It centers on key work processes such as distribution, order entry, or employee training. This type of benchmarking identi¬ fies the most effective practices in companies that perform similar functions, no matter in what industry. For example. Xerox adapted the warehousing and distribution prac¬ tices of L.L. Bean for its spare parts distribution system. Texas Instruments studied the kitting (order preparation) practices of six companies, including Mary Kay Cosmetics, and designed a process that captured the best practices of each of them, cutting kitting cycle time in half. A General Mills plant in Lodi, California, had an average machine changeover time of three hours. Then somebody said, "From three hours to 10 min¬ utes!" Employees went to a NASCAR track and videotaped the pit crews, and studied the process to identify how the principles could be applied to the production changeover processes. Several months later, the average time fell to 17 minutes.53 The U.S. Marine Corps studied companies such as Wal-Mart and United Parcel Service to improve its supply chain processes, changing its inventory policies and learning to employ modern technology like handheld computers. Thus, companies should not aim benchmarking solely at direct competitors or similar organizations; in fact, they would be mistaken to do so. If a company simply benchmarks within its own industry, it may merely be competitive and have a slight edge in those areas in which it is the industry leader. However, if benchmarks are adopted from outside the industry, a company may learn ideas and processes as well as new applications that allow it to surpass the best within its own industry and to achieve distinctive superiority. Finally, strategic benchmarking examines how companies compete and seeks the winning strategies that have led to competitive advantage and market success. The typical benchmarking process can be described by the process used at AT&T.
1. Project conception: Identify the need and decide to benchmark. 2. Planning: Determine the scope and objectives, and develop a benchmarking plan. 3. Preliminary data collection: Collect data on industry companies and similar processes as well as detailed data on your own processes.
351
Part 2
352
Quality in High-Performance Organizations
4. Best-in-class selection: Select companies with best-in-class processes. 5. Best-in-class collection: Collect detailed data from companies with best-in-class processes. ' 6. Assessment: Compare your own and best-in-class processes and develop recom¬ mendations. 7. Implementation planning: Develop operational improvement plans to attain superior performance. 8. Implementation: Enact operational plans and monitor process improvements. 9. Recalibration: Update benchmark findings and assess improvements in processes.54 Reengineering The process of reengineering has been defined as "the fundamental rethinking and radical redesign of business processes to achieve dramatic improve¬ ments in critical, contemporary measures of performance, such as cost, quality, service, and speed."55 Such questioning often uncovers obsolete, erroneous, or inappropriate Reengineering involves asking basic assumptions. Radical redesign involves tossing questions about business processes: out existing procedures and reinventing the Why do we do it? and Why is it process, not just incrementally improving it. The done this way? goal is to achieve quantum leaps in perfor¬ mance. For example, IBM Credit Corporation cut the process of financing IBM computers, software, and services from seven days to four hours by rethinking the process. Originally, the process was designed to handle difficult applications and required four highly trained specialists and a series of handoffs. The actual work took only about 1.5 hours; the rest of the time was spent in transit or delay. By questioning the assumption that every application was unique and difficult to process, IBM Credit Corporation was able to replace the specialists by a single individual supported by a user-friendly computer system that provided access to all the data and tools that the specialists would use. Successful reengineering requires fundamental understanding of processes, cre¬ ative thinking to break away from old traditions and assumptions, and effective use of information technology. PepsiCo has embarked on a program to reengineer all of its key business processes, such as selling and delivery, equipment service and repair, procurement, and financial reporting. In the selling and delivery of its products, for example, customer reps typically experience stockouts of as much as 25 percent of product by the end of the day, resulting in late-day stops not getting full deliveries and the need to return to those accounts. Many other routes return with overstock of other products, increasing handling costs. By redesigning the system to include handheld computers, customer reps can confirm and deliver that day's order and also take a future order for the next delivery to that customer.56 Benchmarking can greatly assist reengineering efforts. Reengineering without benchmarking probably will produce 5 to 10 percent improvements; benchmarking can increase this percentage to 50 or 75 percent. When GTE reengineered eight core processes of its telephone operations, it examined the best practices of some 84 com¬ panies from diverse industries. By studying outside best practices, a company can identify and import new technology, skills, structures, training, and capabilities.57 PROCESS MANAGEMENT IN THE BALDRIGE CRITERIA, ISO 9000, AND SIX SIGMA
Category 6 of the 2003 Malcolm Baldrige National Quality Award Criteria for Perfor¬ mance Excellence is Process Management. Item 6.1, Value Creation Processes, examines
Chapter 7
Process Management
how an organization identifies and manages its key processes for creating customer value and achieving business success and growth. This process includes how an organization incorporates customer and supplier input into determining its key process requirements; how processes are designed to meet these requirements; and how new technology, organizational learning, cycle time, productivity, cost control, and other efficiency and effectiveness factors are designed into processes. This cri¬ teria item also seeks to understand how key performance measures and indicators are used for controlling and improving processes, how costs associated with inspec¬ tions, tests, and audits are minimized, and how defects and rework are prevented. Finally, it calls for information on how value creation processes are improved to achieve better performance, reduce variability, improve products and services, keep processes current with business needs and directions, and how improvements are shared with other organizational units. It might include Six Sigma approaches, use of ISO 9000:2000, or other process improvement tools. Item 6.2, Support Processes, calls for similar information about key support processes, particularly on how they are designed to meet appropriate internal and external customer requirements, and how they are controlled and improved. Many aspects of ISO 9000:2000 deal with process management activities (in fact, the entire standards are focused on an organization's ability to understand, define, and document its processes). For example, one of the requirements is that organiza¬ tions plan and control the design and development of products and manage the inter¬ faces between different groups involved in design and development to ensure effective communication and clear assignment of responsibility. The standards also address the management of inputs and outputs for design and development activi¬ ties, and use of systematic reviews to evaluate the ability to meet requirements, iden¬ tify any problems, and propose necessary actions; purchasing processes; control of production and service, including measurement and process validation; control of monitoring and measuring devices used to evaluate conformity; analysis and improvement; monitoring and measurement of quality management processes; and continual improvement, including preventive and corrective action. The standard requires that an organization use its quality policy, objectives, audit results, data analysis, corrective and preventive actions, and management reviews to continually improve its quality management system's effectiveness. Six Sigma is based on understanding and improving processes on a project-byproject basis. Two of the advantages of Six Sigma are that projects are clearly linked to strategic needs and organizational objectives, and that projects are managed under a common framework. This linkage enables projects to be timely and relevant, and ensures that controls are put in place to leverage the improvements that are identified. The Six Sigma team-project approach provides a natural fit with the requirements of product and process design, control, and improvement. A good system for process management is a prerequisite to Six Sigma. Obviously, to effectively design or improve a process you first need to understand it. If an organization does not have an ongoing system of process management, it will be quite difficult to implement Six Sigma. Some key processes that are necessary to implement Six Sigma include the following: • • • • • •
Project selection and definition Financial review Training Leadership for project leaders Project leader mentoring Certification for Six Sigma specialists
353
Part 2
354
Quality in High-Performance Organizations
• Project tracking and reporting • Information management and dissemination It is important to note that Six Sigma is not a substitute for continuous improve¬ ment. Because of its reliance on specialists—the "black belts" who lead the high-profile projects—it becomes quite easy to ignore simple improvements that can be achieved at the process owner level. In fact, it can easily alienate process owners who, instead of seeking continuous improvements, leave them to the specialists. Thus, the objectives are somewhat different, yet both approaches can easily support one another. Process owners should be trained in Six Sigma methods and be involved in formal Six Sigma projects, but still have responsibility for continuous improvement on a daily basis.
Quality in Practice Gold Star Chili: Process Management58 (We encourage you to read the Gold Star Chili case in Chapter 4 first for background information about the company.) Gold Star Chili, a chain of chili restaurants in the greater Cincinnati area, views process management activities as critical to its business success. Quality improvement teams, technology, and strong relationships with sup¬ pliers ensure that their chili is produced in a con¬ sistent fashion with respect to taste, viscosity, and general quality. Figure 7.7 shows a process-based organization of the company. Three major value-creation processes link the operation of the company to its customers and other stakeholders: 1. Franchising 2. Restaurant operations 3. Manufacturing/distribution Sustaining these processes are various support processes, such as research and development, human resources, accounting, purchasing, opera¬ tions, training, marketing, and customer satisfac¬ tion, as well as design processes for new products, menus, and facilities. Production/delivery processes are coordinated at the corporate office and documented in manuals provided to each store. Internal customer needs are addressed in quality improvement team meetings. The franchising process, outlined in Figure 7.8, is designed to ensure a smooth and successful start-up that meets company objectives. The process has been refined over time and includes extensive interaction with prospective and approved franchisees. New technology has been
introduced to facilitate the process. For example, a site-selection software package is used to evaluate market potential using a variety of demographic data. Computer-aided design is also used for site development. Because franchise process delays are costly, the process helps to eliminate variability, reduce cycle time, and cut down on problems that might occur during development and introduc¬ tion. Procedure manuals have been developed to provide each store with the necessary information and training to ensure that they operate efficiently. Restaurant processes include Cash Register, Steam Table, Drive-Thru, Tables, Bussers, and Management. These processes are designed to ensure that the principal requirements of all cus¬ tomers, such as being served in a timely manner and receiving their order accurately, are met. Prior to the opening of each restaurant, training sessions ensure that these processes are performed correctly and according to company standards. Each employee is cross-trained to perform each function. Chili production is performed at the Gold Star Commissary. A nine-member team, cross-trained to perform each process, is responsible for adding beef, spices, tomatoes, and water during produc¬ tion. The chili must pass a series of strict tests before being shipped to restaurants. Control of chili production is assisted by various pieces of equipment for precise measurement. For example, a Bostwick Viscosity Meter determines the consis¬ tency of the chili, determining whether it is too thick or thin, and a flow meter adds the proper amount of water. Other equipment analyzes the fat content of the ground beef used in the chili.
Chapter 7
Process Management
355
Figure 7.7 Gold Star Chili, Inc. Organization
Source: Courtesy of Gold Star Chili. Used with permission.
The final taste test is performed by members of the commissary to ensure that each batch meets estab¬ lished standards. The commissary team also serves as a quality improvement team. Since July 1996, they have met informally on a daily basis to discuss processes and feedback from internal and external cus¬ tomers; they use a formal improvement process in
reporting their activities. They use information from customer comment cards, measurements of waiting time for drive-through service, and feed¬ back from restaurant managers to analyze and adjust processes as necessary. Store performance and quality are measured quarterly through visits by corporate employees. Monthly meetings of key process leaders and daily team meetings analyze
356
Part 2
Quality in High-Performance Organizations
Figure 7.8 Gold Star Chili Franchising Process
GOLD STAR CHILI, INC. 650 Lunken Park Drive Cincinnati, Ohio 45226 (513)231-4541 Steps to a Gold Star Chili Franchise 1.
Submit a fully completed franchise application. We will respond to you within 15 business days on your applications.
2.
You will receive for your review our Uniform Franchise Offering Circular and Exhibits. At this time you will sign and date the Receipt of Offering Circular.
3.
Ten days after we receive your Receipt of Offering Circular, you will be sent a Confiden¬ tiality Agreement which you must hold for 5 days, then sign, date, and return to us.
4.
Schedule and complete a meeting with support staff of Gold Star Chili in our Cincin¬ nati location.
5.
Attend a two-day orientation in Cincinnati. You will work with our Operations and Training personnel who will review your qualifications and objectives.
6.
You will then be notified of your approval or disapproval of your request to become a Gold Star Chili Franchisee.
7.
Gold Star personnel will begin the process of identifying and approving a restaurant location. A scope of work per Gold Star standards will be completed.
8.
Sign Gold Star Chili Franchise Agreement and pay initial fee.
9.
Begin construction.
10.
Complete the training program.
11.
Develop an opening plan.
12.
Open your Gold Star Chili Restaurant.
Source: Courtesy of Gold Star Chili. Used with permission.
processes for improvement opportunities, such as changes in procedures or the introduction of new technology. For example, several restaurants dis¬ covered large clumps of beef in the chili. The team determined that a new beef pump was not grinding the meat correctly.
The selection of suppliers is driven by two cri¬ teria: quality and price. Gold Star partners with key product suppliers for restaurant equipment and food products. They seek out local companies and educate them in their business needs and practices. For example, they have invited suppliers
Chapter 7
Process Management
to attend a seminar on the Gold Star Chili total quality philosophy and suppliers' role in the process. To ensure that raw materials meet Gold Star specifications, potential and current suppliers visit the commissary to be informed about what Gold Star requires and what technologies the com¬ pany expects them to have. Suppliers are required to meet or exceed quality standards and provide products at reasonable prices. Gold Star recently embarked on establishing a supplier scorecard to measure and monitor supplier performance. The scorecard includes ratings on on-time delivery or service, accuracy of invoicing and shipping docu¬ ments, customer service, cost and value, technical expertise, and supplier quality initiatives, and seeks to identify strengths and targeted areas for improvement in each key area. Gold Star attempts to establish long-term rela¬ tionships with its suppliers. Company managers visit suppliers' facilities on a regular basis to solicit comments and complaints, and to discuss areas for improvement. During these discussions, suppliers often provide Gold Star with information about new technologies, suggestions for process improvements, and other helpful knowledge. For
357 example, by sharing information with one key paper supplier, the supplier was able to redesign Gold Star's purchasing process, enabling the sup¬ plier to increase minimum order levels for deliv¬ eries, which reduced Gold Star's overall costs. The company conducts annual cost audits to deter¬ mine whether costs might be lowered without sac¬ rificing quality. If an alternative supplier is found with similar quality, service, and lower costs. Gold Star will approach its current supplier with the opportunity to lower costs. Key Issues for Discussion
1. How does the organization structure in Figure 7.7 reflect Deming's view of a produc¬ tion system as discussed in Chapter 1? 2. As a small, privately held company, Gold Star is relatively new at applying total quality management approaches to its process management. Based on the informa¬ tion provided here, what suggestions might you provide in the process management area as the company matures in its journey to total quality?
Quality in Practice Bringing Process Management to Education59 In 1991, 200 angry parents, mostly Hispanic, con¬ fronted Gerald Anderson in his first week as super¬ intendent of the Brazosport Independent School District (BISD), then considered among the worst districts in Texas. They demanded to know why their children had the worst test scores and what he was going to do about it. He seized the chal¬ lenge, developing process-based techniques that raised achievement levels of all students. Today, this 13,500-student school district 50 miles south of Houston is the largest Exemplary school district in the state of Texas, a designation earned by only 121 districts based on tough accountability ratings. Bra¬ zosport was one of only two educational institu¬ tions to receive site visits for the Malcolm Baldrige National Quality Award in the first year of educa¬ tion sector eligibility (1999). Schools across the country have begun to benchmark their processes. Brazosport is particularly noteworthy because
of the socioeconomic diversity of its students. The success of BISD has been a never-ending journey of dedication and hard work, based on a philos¬ ophy of no excuses, and the belief that all children can learn regardless of family background, sex, or socioeconomic status. The strategy that Brazosport employed was an eight-step process to improve the learning process. The process, based on TQ principles and Effective Schools research that iden¬ tifies characteristics of schools where all students succeed, is summarized here. 1. Educators examine results of state profi¬ ciency tests, identifying areas where students need to improve. 2. Teachers develop a time line, determining what they will teach and how much time they'll spend on each objective based on the needs of the students.
358
Part 2
3. Teachers devise daily 10-minute segments to work on concepts where students need help. Each teacher gets an instructional focus sheet stating the objectives to be taught, dates for teaching each objective, and dates when stu¬ dents will be assessed on them. 4. Students are constantly assessed to deter¬ mine whether they have mastered the con¬ cepts, giving teachers new data. 5. Students receive tutorial time so that teachers can reteach areas that students have not mastered. 6. Students who master the concepts take part in enrichment activities. 7. Teachers receive maintenance booklets to help them reteach key concepts and keep stu¬ dents on track. They meet frequently in teams to review progress. 8. Principals monitor the instructional process by visiting classrooms and meeting with teachers to discuss students' progress.
Quality in High-Performance Organizations
"The process is nothing but effective teaching practices," states Patricia Davenport, former director of curriculum and instruction at Bra¬ zosport. However, its uniqueness lies in its imple¬ mentation as a disciplined process. Brazosport piloted the program for two years in Velasco Ele¬ mentary, its school with the highest percentage of economically disadvantaged students—82 per¬ cent—and the lowest test scores on the state assessment. In two years' time, students went from less than 30 percent mastering the state assessment to 70 percent. Figure 7.9 shows the results of standardized mathematics tests over an eight-year period. Sim¬ ilar results were achieved in writing and reading as well. The American Productivity and Quality Center, a Houston-based consulting firm, trains school districts across the nation in this process.
Figure 7.9 Mathematics Standardized Test Results for Brazosport Independent School District
All Students
—O— A. American
—Hispanic
—X— White
—I—
Economically Disadvantaged
Source: Brazosport Independent School District, Freeport, Texas.
Chapter 7
Process Management
Key Issues for Discussion
1. How does the eight-step process exemplify principles of total quality and process man¬ agement?
359 2. What infrastructure would a school district need to make this process effective? Com¬ ment specifically on leadership and human resource issues that would need to be addressed.
Review Questions 1. Define process management and its key components. Why is it important to any business?
2. Summarize the principles on which AT&T bases its process management methodology. Define and illustrate the principal categories of processes. Why must processes be repeatable and measurable? Summarize the leading practices in process management. Describe the product design and development process. How can product design affect manufacturability? Explain the concept and importance of design for manufacturability. 8. Summarize the key design practices for high quality in manufacturing and assembly. 9. Discuss social responsibility issues relating to product design facing businesses today. 3. 4. 5. 6. 7.
10. Discuss the importance of and impediments to reducing the time for product development.
11. Describe the basic approach used for designing value-creation and support processes.
12. Explain the differences between designing manufactured products and ser¬ vices. How should the design of services be approached? 13. Describe the three components of any control system. 14. How can one check whether process owners have true responsibility for con¬ trolling a process? 15. Explain the concept of after-action review. 16. Why is it important to establish strong relationships with suppliers? What are some good supplier management practices? (See Bonus Materials.) 17. What is the purpose of supplier certification? Explain some of the common practices for supplier certification. (See Bonus Materials.) 18. Explain the Japanese concept of kaizen. How does it differ from traditional Western approaches to improvement? 19. How are projects considered as vital value-creation processes? 20. Explain the life cycle of a project from a TQ perspective. 21. What aspects of Six Sigma projects are process-related? Briefly define them. 22. What is flexibility and why is it important to a modern organization? 23. What are the key impacts of cycle time reduction? 24. What is a stretch goal? How can stretch goals help an organization? 25. Define benchmarking and list its benefits. 26. What is reengineering? How does it relate to TQ practices? 27. Discuss how process management is addressed in the Baldrige criteria, ISO 9000:2000, and Six Sigma.
360
sr
Part 2
Quality in High-Performance Organizations
Discussion Questions 1. Identify some of the key processes associated with the following business activ¬ ities for a typical company: sales and marketing, supply chain management, managing information technology, and managing human resources. 2. Provide some examples of processes that are repeatable and measurable and some that are not. 3. List some of the common processes that a student performs. How can these processes be improved? 4. Are classroom examinations a means of control or improvement? What should they be? 5. Why are modern products more difficult to manufacture than traditional prod¬ ucts such as bicycles or hand tools? 6. How can kaizen be applied in a classroom? 7. The kaizen philosophy seeks to encourage suggestions, not to find excuses for failing to improve. Typical excuses are "If it's not broken, don't fix it," "I'm too busy to work on it," and "It's not in the budget." Think of at least five other excuses why people don't try to improve. 8. How might Six Sigma projects be applied to course design? 9. What is the "product development process" a school might use for designing and introducing a new course? How might it be improved to reduce "time-to-market"? 10. How can a manager effectively balance the three key components of a service system design? 11. In a true story related by our colleague Professor James W. Dean, Jr., the general manager of an elevator company was frustrated with the lack of cooperation between the mechanical engineers who designed new elevators and the manu¬ facturing engineers who determined how to produce them.60 The mechanical engineers would often completely design a new elevator without consulting with the manufacturing engineers, and then expect the factory to somehow figure out how to build it. Often the new products were difficult or nearly impossible to build, and their quality and cost suffered as a result. The designs were sent back to the mechanical engineers (often more than once) for engi¬ neering changes to improve their manufacturability, and customers sometimes waited for months for deliveries. The general manager believed that if the two groups of engineers would communicate early in the design process, many of the problems would be solved. At his wits' end, he found a large empty room in the plant and had both groups moved into it. The manager relaxed a bit, but a few weeks later he returned to a surprise. The two groups of engineers had finally learned to cooperate—by building a wall of bookcases and file cabinets right down the middle of the room, separating them from each other! What would you do in this situation? 12. Legal Sea Foods operates several restaurants and fish markets in the Boston area and other East Coast locations. The company's standards of excellence mandate that it serves only the freshest, highest-quality seafood. It guarantees the quality by buying only the "top of the catch" fish daily. Although Legal Sea Foods tries to make available the widest variety every day, certain species of fish are subject to migratory patterns and are not always present in New England waters. Weather conditions may also prevent local fishermen from fishing in certain areas. Freshly caught fish are rushed to the company's quality control center where they are cut and filleted in an environmentally controlled state-of-the-art
Chapter 7
Process Management
facility. All shellfish come from government-certified beds and are tested in an in-house microbiology laboratory for wholesomeness and purity. There are even special lobster storage tanks so that all lobsters are held under optimum conditions, in clean, pollution-free water. Every seafood item is inspected for quality eight separate times before it reaches the table. At Legal Sea Foods' restaurants, each meal is cooked to order. Even though servers make every effort to deliver all meals within minutes of each other, they will not jeopardize the quality of an item by holding it beneath a heat lamp until the entire order is ready. The service staff is trained to work as a team for better service. More than one service person frequently delivers food to a table. When any item is ready, the closest available person serves it. Customer questions can be directed to any employee, not just the person who took the initial order. a. What are the major processes performed by Legal Sea Foods? How does the process design support its goal of serving only the freshest, highest-quality seafood? b. Where would Legal Sea Foods fall on the three-dimensional classification of service organizations? Is its process design consistent with this classification? 13. The president of Circle H assigned you to perform a complete investigation to determine the causes of certain quality problems and to recommend appro¬ priate corrective action. You have authority to talk to any other person within the company. The early stages of your investigation establish that the three reasons most often cited by customers are symptomatic of some major quality problems in the company's operations. In proceeding with the audit, you decide to review all available data, which may yield indications of the root causes of these problems. Further investigation reveals that, over a recent four-month period, a proce¬ dural change was made in the order approval process. You wish to find out whether this change caused a significant difference in the amount of time required to process an order from field sales through shipping. You therefore decide to investigate this particular situation. On completion of your investigation into the problems with order pro¬ cessing, you determine that the change in procedures for order approval led to an increase in the amount of time required to restock goods in the customers' stores. You want to recommend corrective action for this problem, but you first do additional investigation as to why the change was made. You learn that, because of large losses on delinquent accounts receivable, the change was made to require that the credit manager approve all restock orders. This approval requirement added an average of three hours to the amount of internal pro¬ cessing time needed for a restock order. On review of your report, the president of Circle H takes note of administra¬ tive problems whose existence he had never suspected. To assure that corrective action will be effective and sustained, the president assigns you to take charge of the corrective action program.1'1 a. What types of data would be most useful to review for clues as to why the three major customer complaints occurred? b. How would you investigate whether the change in the order approval process had a significant effect on order processing time? c. Given your knowledge of problems in both order processing and accounts receivable, what should you do? 14. McDonald's used to make food to stock, storing sandwiches in a large tray used to fulfill customer orders. When sales went flat in the mid 1990s and independent
361
362
Part 2
Quality in High-Performance Organizations
market testing showed a widening gap with competition in food quality, McDonald's recognized that the make-to-stock process was not meeting customer demands. After five years of lab and market testing, McDonald's rolled out the new "Just for You" system, which began in March 1998, to create a make-to-order environment. This shift required a massive change in technology with computers to coordinate orders; food production equipment using "rapid toasters" and tem¬ perature-controlled "launching zones" to replace the old heat lamps and holding bins; new food preparation tables, and retraining efforts for the entire domestic food production organization of more than 600,000 crew members. However, this system has apparently backfired. Sales did not improve as expected and cus¬ tomers complained about slow service. The new system increased the average ser¬ vice time 2 to 3 minutes per order, and 15-minute waits were not uncommon. McDonald's stock price decreased, and rivals such as Wendy's captured addi¬ tional market share.62 What lessons does this experience suggest for process man¬ agement? What might McDonald's have done differently? 15. The Cincinnati Water Works (CWW) serves approximately 1 million customers.63 Its billing system allows customer service representatives (CSRs) to retrieve information from customer accounts quickly using almost any piece of data such as customer name, address, phone number, social security number, and so on. Besides a customer's account history, the system contains everything that was said in a call, including documentation of past problems and their resolution. An integrated voice response system provides automated phone support for bill paying and account balances, tells customers of the approximate wait time to speak to a CSR, and allows the customers to leave a message for a CSR to return a call. An information board in the department shows the number of customers waiting, average length of time waiting, and the number of CSRs that are busy and doing post-call work. A pop-up screen provides CSRs with customer data before the phone rings so that he or she will have the customer's information before they even say hello. Work orders taken by CSRs, such as a broken water main or leaking meter, are routed automatically to a field service supervisor for immediate attention. This system is also used internally to allocate maintenance workers when a problem arises at a pumping station or treatment facility. A geo¬ graphic information system is used for mapping the locations of water mains and fire hydrants, and provides field service employees, meter readers, and con¬ tractors exact information to accomplish their work. Handheld meter readers are used to locate meters and download data into computers. Touch pad devices provide exterior connections to inside meters, eliminating the necessity to enter a house or building. CWW is also investigating automated meter readers and radio frequency devices that simply require a company van to drive by the building to automatically obtain readings. Discuss how technology has affected the processes of CWW. What specific types of improvements (quality, cycle time, etc.) were these applications designed to address? Can you think of similar uses of these technologies in other service applications?
jj§jj|
Projects, Etc. 1. Identify some of the major processes a student encounters in a college or uni¬ versity. What types of noneducational institutions perform similar processes and might be candidates for benchmarking?
Chapter 7
Process Management
363
2. Write down your process for preparing for an exam. How could this process be improved to make it shorter and/or more effective? Compare your process to those of your classmates. How might you collectively develop an improved process? 3. Interview a plant manager at a local factory to determine his or her philosophy on process management. What techniques does the company use? 4. Investigate design-for-environment practices in some of your local industries. Describe company policies and the methods and techniques that they use to address environmental concerns in product design. 5- Christina Clark works at a food service operation for a large amusement park. She has been charged with developing a process control plan based on HACCP principles for meeting food safety requirements (see the Bonus Materials). For example, the requirements for hot dogs include: • Receiving: Refrigerated hot dogs should be between -40 and 34 degrees Fahrenheit when received. • Storage: Storage temperature should be between -40 and 34 degrees Fahren¬ heit. • Cooking: Hot dogs should be heated to a temperature of 145 ± 5°F within 30 minutes of placing on the grill. • Cooked Storage: Leftover hot dogs must be covered and placed in refrigeration immediately and reach a temperature of 40°F or lower within 4 hours. • Reheating: Hot dogs must be reheated to an internal temperature of 165 ± 5°F within 20 minutes, one time only. Develop a process control plan for ensuring that these requirements are met. Design any forms or "standard operating procedures" that you think would be helpful in implementing your plan. 6. Design a process for the following activities: a. Preparing for an exam b. Writing a term paper c. Planning a vacation d. Making breakfast for your family e. Washing your car Draw a flowchart for each process and discuss how ways in which both quality and cycle time might be improved. 7. Design an instrument for evaluating the "process orientation" of an organiza¬ tion. For example, what characteristics would you look for in firms that have a strong process orientation?
Additional cases, including Baldrige assessment cases, are available in the Bonus Materials Folder on the CD-ROM. I. The State University Experience
Wow! That State University video was really cool.
like to be a student at State. Hmmm, I think I'll ask
It has lots of majors; it's close to home so I can
Mom and Dad to take a campus tour with me. . . .
keep my job; and Mom and Dad loved it when they visited. I wish I could know what it's really
I'm sure that we took our tour on the hottest day of the summer. The campus is huge—it took
Part 2
364
Quality in High-Performance Organizations
us about two hours to complete the tour and we
send in the confirmation form. It really looks a lot
didn't even see everything! I wasn't sure that the
like the application. In fact, I know I gave them a
tour guide knew what he was doing. We went into
lot of the same information. I wonder why they
a gigantic lecture hall and the lights weren't even
need it again? Seems like a waste of time. . . . Orientation was a lot of fun. I'm glad they
on. Our tour guide couldn't find them so we had to hold the doors open so the sunlight could come
straightened out my acceptance at U. College. I
in. About three-fourths of the way through the
think I will enjoy State after all. I met lots of other
tour, our guide said, "State University isn't really a
students. I saw my advisor and I signed up for
bad place to go to school; you just have to learn
classes. All I have left to do is pay my tuition bill.
the system." I wonder what he meant by that? . . .
Whoops. None of my financial aid is on this bill. I
This application is really confusing. How do I
know I filled out all of the forms because I got an
let the admissions office know that I am interested
award letter from State. There is no way my par¬
in physics, mechanical engineering, and industrial
ents and I can pay for this without financial aid. It
design? Even my parents can't figure it out. I
says at the bottom. I'll lose all of my classes if I
guess I'll call the admissions office for some
don't pay the bill on time. . . .
help. . . . I'm so excited! Mom just handed me a letter
my form and the fee a long time ago. What am I
from State! Maybe they've already accepted me.
going to do? I don't want to lose all of my classes.
What? What's this? They say I need to send my
I have to go to the admissions office or my college
transcript. I did that when I mailed in my applica¬
office and get a letter that says I am a confirmed
tion two weeks ago. What's going on? I hope it
student. O.K. If I do that tomorrow, will I still have
won't affect my application. I'd better check with
all of my classes? . . .
I'm not confirmed on the computer? I sent in
I can't sleep; I'm so nervous about my first
Admissions. . . . You can't find my file? I thought you were only
day. . ..
missing my transcript. I asked my counselor if she had sent it in yet. She told me that she sent it last
Discussion Questions
week. Oh, you'll call me back when you locate my
1. What breakdowns in service processes has
file? O.K. ...
this student experienced?
Finally, I've been accepted! Wait a minute. I
2. What types of process management activities
didn't apply to University College; that's a two-
should State University administrators
year program. I wanted physics, M.E., or indus¬
undertake?
trial design. Well, since my only choice is U. College and I really want to go to State, I guess I'll
II. The PIVOT Initiative at Midwest Bank, Part I64 Midwest, a bank holding company, is located in
services 24-hours per day via its network of ATMs,
Ohio. Its main subsidiary provides a diverse line
a 24-hour telephone customer service center, or
of banking and financial products and services
online. This bank has served the financial needs of
regionally; and selected business activities are con¬
its customers for 100 years, and currently 3,200
ducted nationally. Consumer, small business, and
associates serve approximately 600,000 customers.
investment products and services are offered through a network of retail banking centers
The PIVOT Initiative
located primarily within Ohio and Kentucky. Mid¬ west Bank also has a growing presence in Florida
The PIVOT initiative, the name that Midwest gave
with 13 retail banking centers. Commercial
its Six Sigma process improvement approach,
banking products and services are offered through
started with the selection of three pilot projects,
nine regional offices. Customers can also access
one of which was in the Commercial Processing
this Midwest's financial and banking products and
Department (CPD) that works as Midwest Bank's
Chapter 7
Process Management
cash vault. CPD already operated at a high level of sigma (4.26) as found early in Yellow Belt Six Sigma training. CPD processes a high dollar volume of transactions. One costly error in the previous year resulted in a loss of over a quarter million dollars and brought the department to the forefront of change initiatives. Once the project was chosen, the bank selected six associates to run the first PIVOT project. A project coordinator working from the project office was selected as project manager for the Six Sigma functions of the project. An operations financial manager was in charge of financial impact analysis and equipment purchasing. The assistant vice president and team supervisor, were subject matter experts from within CPD. Another project coordinator was brought on board for her bank wide knowledge and overall project support. The project analyst for CPD and five other areas was in charge of the departmental project management and was the Six Sigma analyst for the team. The Six Sigma analyst was responsible for data integrity, graphical analysis, and data stratification. After their weeklong Six Sigma course, the six team members ran the project from project definition to the control stage. The team followed the Six Sigma DMAIC process steps (Define, Measure, Analyze, Improve, Control) during the CPD PIVOT project to define the project and get it underway. DMAIC Define Stage
The senior vice president and vice president over CPD were the champions for this project and ini¬ tially worked to establish the problem definition statement. These champions were responsible for the process every day and also held accountable for the errors in the department. Because the two largest potential sources of errors (strapping and deposit processing) did not influence each other in the process and had sepa¬ rate causes for creating errors, the champions sep¬ arated them. The problem statement defined the number of errors the department was accountable for during the previous year and the dollar losses of these errors. In this case study, the number of errors and the actual dollar losses are only approx¬ imate. The (disguised) problem statement was: In (the previous year) the number of internal and external defects for the CPD
365 was 150, resulting in Bank losses of $400,000 as well as significant potential risk exposure. Included in the losses is an anomaly of $280,000. The remainder rep¬ resents a gap of $120,000 versus the goal of $0 of total losses due to Commercial Processing Department operations. Our objective is to reduce the internal error ratio by December of the current year and the total amount of losses by over 50% in the following 12 months. Projects from this business case will reduce loss expense and risk exposure, while increasing customer satisfaction. Much debate centered on whether to include the anomaly loss since it skewed the numbers con¬ siderably. However, the decision was finally made to include it. Support for the CPD PIVOT project centered on risk mitigation, which is difficult to quantify, and on dollar losses required to carry the project. Based on the potential reduction of approximately $400,000 in losses and the future risk mitigation, the steering committee approved the project launch, and the CPD PIVOT team moved onto the Measure stage. (See "The PIVOT Initiative at Midwest Bank, Part II" in Chapter 10 for a continuation of this case.) Discussion Questions
1. What conclusions can you reach on the importance of team preparation and member selection to the "Define" stage, and eventual success of Six Sigma projects, such as the PIVOT project? 2. How do the roles of the PIVOT team mem¬ bers, described in the case, match or not match the roles in Table 7.4 in the chapter? Why do you think that they differ? 3. What factors do you think weighed in the decision to include the $280,000 "anomaly" in the project justification? If you were the project champion, how would you assess this justification in deciding on whether the pro¬ ject was significant enough to move forward?
366
Part 2
Quality in High-Performance Organizations
III. Stuart Injection Molding Company65
Stuart Injection Molding Company (SIMC) is a small business that specializes in custom plastic molding for many different industries, including appliances and various consumer products such as toys. Adele Stuart, daughter of the company's founder and current CEO wishes to expand into the automotive sector. However, she realizes that to do so will require more formalized systems and eventually ISO 9000 registration. Although many basic procedures for assuring quality are in place, most have been conducted informally, and the company never compiled a formal quality manual that documents the system and outlines specific responsibilities for managers and workers. Recog¬ nizing the lack of a manual as a major deficiency, they called you in as a consultant to help. After spending some time in the plant talking with many employees, you jotted down several notes and observations: • The plant manager (PM) is responsible for ensuring the success of the quality manage¬ ment system by providing the necessary resources and reviewing system perfor¬ mance. However, SIMC has a quality assur¬ ance (QA) department that is responsible for the majority of implementation issues, such as maintaining measuring and test equip¬ ment, verifying process capability, per¬ forming inspection, selecting methods for monitoring process performance, and auditing the system. • All functional departments recognize their responsibility for quality planning and pro¬ ducing high-quality products. For example, the marketing and sales department con¬ ducts market research to understand cus¬ tomer needs and handles customer complaints; the project engineering depart¬ ment performs design reviews; the manufac¬ turing department conducts in-process inspection for the purpose of maintaining control and coordinates continuous improve¬ ment processes. Maintenance, supplier rela¬ tions, receiving, and human resources departments support these functions. • Most products are custom-designed with the customer. When a new job is contracted, a
cross-functional team is selected that includes members from project engineering, quality assurance, manufacturing, and sales. This team develops all the specifications to ensure that design meets customer require¬ ments and can be made according to these requirements, selects materials and process tolerances, determines production routings and inspection plans, develops a production control plan and measurement system, and monitors a trial production run. The cus¬ tomer must approve all design changes. The company uses a variety of contemporary tools to simplify and optimize the product while also focusing on reducing production cost and waste. These tools include quality function deployment, geometric dimensioning and tolerancing, design for manufacturing and assembly, value engineering, design of experi¬ ments, failure mode and effects analysis, and cost/performance/risk analysis. Inspection is routine during the production process. QA lab personnel perform all phases of inspection and testing. Production opera¬ tors use a "first-piece" inspection process to validate the start-up for a new product. Receiving inspection is performed on mate¬ rial, purchased parts, and subassemblies used in processing, manufacturing, and assembly. Operators also perform in-process inspections during production and final inspection on fin¬ ished products. When contractually required, statistical process control techniques are used to ensure control of key process characteris¬ tics. Gauging instruction sheets are main¬ tained by QA for at least one year. Nonconforming products are labeled with a "Do Not Use" tag and kept from being shipped. This tag describes the nonconfor¬ mance, documents the disposition decision, and records the reinspection results. If they are repaired or reworked, they are rein¬ spected. Products that do not fully comply with specific requirements are not shipped without customer authorization. When non¬ conformities are detected, a cross-functional team investigates them and corrective actions are initiated to prevent their recurrence.
Chapter 7
Process Management
SIMC has a continuous improvement philosophy that pervades the entire organi¬ zation. Processes are improved beyond minimum requirements when further improvements benefit customers. Quality performance and productivity are continu¬ ously monitored to identify opportunities for improvement. Everyone in the organization is encouraged to come forward with ideas for improving products, processes, systems, and productivity within the working environ¬
367 ment. Some examples of opportunities for quality and productivity improvement are the reduction of cycle times, less scrap, rework and repair rates, less unscheduled machine downtime, and process perfor¬ mance variation. Based on this information, what would you rec¬ ommend to the company? Specifically on their process management activities, note any addi¬ tional information that you might need.
IV. CapStar Health Systems: Process Management
The complete CapStar case study, a fictitious example of a Baldrige application, can be found on the CD-ROM that accompanies this book. If you have not read the organizational profile yet (see Case III in Chapter 3), please do so first. Examine their response to Category 6 in the context of the leading practices described in this chapter. (You
need not consider the actual Baldrige criteria for this activity.) What are their strengths? What are their weaknesses and opportunities for improve¬ ment? What specific advice, including useful tools and techniques that might help them, would you suggest?
ENDNOTES 1. Adapted from Katrina Brooker, "The Nightmare Before Christmas," Fortune, January 24, 2000, 24-25. 2. Robert Hof, Debra Sparks, Ellen Neubome, and Wendy Zellner, "Can Amazon Make It?" Business Week, July 10, 2000, 38-43. 3. A. Blanton Godfrey, "Planned Failures," Quality Digest, March 2000, 16. 4. AT&T Quality Steering Committee, Process Quality Management & Improvement Guidelines, AT&T Publication Center, AT&T Bell Laboratories (1987). 5. Justin Martin, "Are You As Good As You Think You Are?" Fortune, September 30,1996,142-152. 6. Michael Hammer and James Champy, Reengi¬ neering the Corporation (New York: HarperBusiness, 1993), 177-178. 7. Steven H. Wildstrom, "Price Wars Power Up Quality," Business Week, September 18,1995, 26. 8. Philip A. Himmelfarb, "Fast New-Product Devel¬ opment at Service Sector Companies," Quality Digest, April 1996, 41-44. 9. Justin Martin, "Ignore Your Customer," Fortune, May 1,1995,121-126. 10. Wolfgang Schneider, "Test Drive Into the Future," BMW Magazine 2,1997, 74-77. 11. Peter J. Kolesar, "What Deming Told the Japanese in 1950," Quality Management Journal 1, no. 1 (Fall 1994), 9-24.
12. Ames Rubber Corporation, Application Sum¬ mary for the 1993 Malcolm Baldrige National Quality Award. 13. Adapted from Douglas Daetz, "The Effect of Product Design on Product Quality and Product Cost," Quality Progress, June 1987, 63-67. © 1987, HewlettPackard Co. All rights reserved. Reprinted with permis¬ sion. 14. Carolyn Lochhead, "Liability's Creative Clamp Holds Firms to the Status Quo," Insight, August 29, 1988, 38-40. 15. John H. Farrow, "Product Liability Require¬ ments," Quality Progress, May 1980, 34-36; Mick Birm¬ ingham, "Product Liability: An Issue for Quality," Quality, February 1983, 41-42. 16. Randall Goodden, "Quality and Product Lia¬ bility," Quality Digest, October 1995, 35-41. 17. David Pescovitz, "Dumping Old Computers— Please Dispose of Properly," Scientific American 282, no. 2 (February 2000), 29; http://www.sciam.com/2000/ 0200issue/0200techbus2.html. 18. Early discussions of this topic can be found in Bruce Nussbaum and John Templeton, "Built to Last— Until It's Time to Take It Apart," Business Week, Sep¬ tember 17,1990, 102-106. A more recent reference is Michael Lenox, Andrew King, and John Ehrenfeld, "An Assessment of Design-for-Environment Practices in
368 Leading U.S. Electronics Firms," Interfaces 30, no. 3 (May/June 2000), 83-94. 19. Nussbaum and Templeton (see note 18). 20. Charles Huber and Robert Launsby, "Straight Talk on DFSS," Six Sigma Forum Magazine, August 2002, 21. 21. Valerie Reitman and Robert L. Simison, "Japanese Car Makers Speed Up Car Making," The Wall Street Journal (December 29,1995), 17. 22. Don Clausing and Bruce H. Simpson, "Quality by Design," Quality Progress, January 1990,41M4. 23. For the fascinating story of how Chrysler redesigned itself, along with their design process, see Brock Yates, The Critical Path (Boston: Little, Brown and Co., 1996). 24. "A Smarter Way to Manufacture," Business Week, April 30,1990. 25. Kelly Scott, "How Federal Express Delivers Cus¬ tomer Service," APICS—The Performance Advantage, November 1999, 44-46. 26. Rebecca Duray and Glenn W. Milligan, "Improving Customers Satisfaction Through Mass Cus¬ tomization," Quality Progress, August 1999, 60-66. 27. Sarah Anne Wright, "Putting Fast-Food to the Test," The Cincinnati Enquirer, July 9, 2000, FI, 2. 28. John Haywood-Farmer, "A Conceptual Model of Service Quality," International Journal of Operations and Production Management 8, no. 6 (1988), 19-29. 29. Charles D. Zimmerman, III, and John W. Enell, "Service Industries," Sec. 33 in J. M. Juran (ed.), Juran's Quality Control Handbook, 4th ed. (New York: McGrawHill, 1988). 30. Paula K. Martin and Karen Tate, "Projects That Get Quality Treatment," The Journal for Quality and Par¬ ticipation, November/December 1998, 58-61. 31. Custom Research Incorporated, "Highlights of CRTs Best Practices," 1996 Baldrige Application Abstract, 13-14. 32. Timothy J. Kloppenborg and Joseph A. Petrick, Managing Project Quality (Vienna, VA: Management Concepts, 2003), 9, 11. 33. "Coca-Cola: A Taste for Quality," The Coca-Cola Company, Atlanta, Georgia. 34. http://www.thecoca-colacompany.com/news/ NewsDetail.asp?NewsDate=6/15 / 99. 35. "Testing for Conformity: An Inside Job," Golf Journal, May 1998, 20-25. 36. Douglas H. Harris and Frederick B. Chaney, Human Factors in Quality Assurance (New York: John Wiley & Sons, Inc., 1969). 37. "DaimlerChrysler's Quality Practices Pay Off for PT Cruiser," News and Analysis, Metrologyworld.com, (accessed March 23, 2000). 38. Adapted from The Ritz-Carlton Hotel Company, Application Summaries for the Malcolm Baldrige National Quality Award, 1992 and 1999.
Part 2
Quality in High-Performance Organizations
39. Adapted from K. M. Casarreal, J. I. Mills, and M. A. Plant, "Improving Service Through Patient Surveys in a Multihospital Organization," Hospital & Health Ser¬ vices Administration, Health Administration Press, Ann Arbor, MI (March/April 1986), 41-52. © 1986, Founda¬ tion of the American College of Health Care Executives. 40. Robert A. Gardner, "Resolving the Process Paradox," Quality Progress, March 2001, 51-59. 41. Andrew E. Serwer, "Michael Dell Turns the PC World Inside Out," Fortune, September 8, 1997, 76-86. 42. Bill Gates with Collins Hemingway, Business @ the Speed of Thought (New York: Warner Books, 1999). 43. Masaaki Imai, KAIZEN—The Key to Japan's Com¬ petitive Success (New York: McGraw-Hill, 1986). 44. Alan Robinson (ed.), Continuous Improvement in Operations (Cambridge, MA: Productivity Press, 1991). 45. Lea A. P. Tonkin, "Kaizen BlitzSM 5: BottleneckBashing comes to Rochester, NY," Target 12, no. 4 (September-October 1996), 41-43. 46. Mark Oakeson, "Makes Dollars & Sense for Mer¬ cedes-Benz in Brazil," HE Solutions (April 1997), 32-35. 47. David A. McCamey, Robert W. Bogs, and Linda M. Bayuk, "More, Better, Faster From Total Quality Effort," Quality Progress, August 1999,43-50. 48. Lawrence S. Pryor, "Benchmarking: A SelfImprovement Strategy," Journal of Business Strategy, November/December 1989, 28-32. 49. Robert C. Camp, Benchmarking: The Search for Industry Best Practices That Lead to Superior Performance (Milwaukee. WI: ASQC Quality Press and UNIPUB/Quality Resources, 1989). 50. Shawn Tully, "Why to Go for Stretch Targets," Fortune, November 14,1994,45-58. 51. Christopher E. Bogan and Michael J. English, "Benchmarking for Best Practices: Winning Through Innovative Adaptation," Quality Digest, August 1994, 52-62. 52. Cathy Hill, "Benchmarking and Best Practices," The 54th Annual Quality Congress Proceedings of the American Society for Quality, 2000. 53. John Hackl, "New Beginnings: Change is Here to Stay," Quality Progress, February 1998, 5. 54. AT&T Consumer Communication Services Sum¬ mary of 1994 Application for the Malcolm Baldrige National Quality Award. 55. Hammer and Champy (see note 6). 56. P. Kay Coleman, "Reengineering Pepsi's Road to the 'Right Side Up' Company," Insights Quarterly 5, no. 3 (Winter 1993), 18-35. 57. Bogan and English (see note 51). 58. We wish to thank Andy Assaley, Scott Atkinson, Frank Cornell, and Eugene Wulsin for their work on which this case is based. Courtesy Gold Star Chili, Inc. 59. Adapted from Cindy Kranz, "Schools Intend to Improve," Cincinnati Enquirer, October 15,2001, pp. Bl, B5.
Chapter 7
Process Management
60. James R. Evans and James W. Dean, Jr. Total Quality: Management, Organization, and Strategy, 3rd ed., Cincinnati, OH: South-Western College Publishing, 2003. 61. Adapted from ASQ Quality Auditor Certification Brochure, July 1989. 62. John E. Ettlie, "What the Auto Industry Can Learn from McDonald's," Automotive Manufacturing & Production, October 1999, 42; David Stires, "Fallen Arches," Fortune, April 29, 2002, 74-76. 63. Adapted from a student project by one of the author's students, Tim Planitz, December 2001.
369 64. Appreciation is expressed to one of the author's students, Michael Wolf, who wrote the paper on which this case is based, as part of the requirements for MGT 699, Total Quality Management, 2002, at Northern Ken¬ tucky University, and Cathy Ernst, senior vice president at the bank. 65. We thank our former students Nick Dattilo, Brian Kessler at Woodcraft Pattern Works, Inc., and Tameka Flowers, on whose research this case is based.
BIBLIOGRAPHY Ahmed, Pervaiz K. and Mohammed Rafiq. "Inte¬ grated Benchmarking: A Holistic Examination of Select Techniques for Benchmarking Analysis," Benchmarking for Quality Management & Technology 5, no. 3 (1998), 225-242. Andersen, Bjorn. Business Process Improvement Toolbox. Milwaukee, WI: ASQ Quality Press, 1999. AT&T Quality Steering Committee. Batting 1000. AT&T Bell Laboratories, 1992. -. Process Quality Management & Improvement Guidelines. AT&T Bell Laboratories, 1987. Boser, Robert B., and Cheryl L. Christ. "Whys, Whens, and Hows of Conducting a Process Capability Study." Presentation at the ASQC/ASA 35th Annual Fall Technical Conference, Lexington, Kentucky, 1991. Brassard, Michael and Diane Ritter. The Memory Jogger II. Methuen, MA: GOAL/QPC, 1994. Burke, Charles J. "10 Steps to Best-Practices Bench¬ marking," Quality Digest (February 1996), 23-28. Camp, Robert C. Business Process Benchmarking: Finding and Implementing Best Practices. Milwaukee, WI: ASQC Quality Press, 1995. DeToro, Irving, and Thomas McCabe. "How to Stay Flexible and Elude Fads," Quality Progress, March 1997, 55-60. Duncan, Acheson J. Quality Control and Industrial Statistics, 5th ed. Homewood, IL: Richard D. Irwin, 1986. Gitlow, H., S. Gitlow, A. Oppenheim, and R. Oppenheim. Tools and Methods for the Improvement of Quality. Homewood, IL: Irwin, 1989. Goetsch, David L. and Stanley B. Davis. Under¬ standing and Implementing ISO 9000:2000. Upper Saddle River, NJ: Prentice Hall, 2002. Godfrey, Blan. "Future Trends: Expansion of Quality Management Concepts, Methods and Tools to All Indus¬ tries," Quality Observer 6, no. 9 (September 1997), 40M3, 46.
Hurley, Heather. "Cycle-Time Reduction: Your Key to a Better Bottom Line," Quality Digest (April 1996), 28-32. Lapin, Lawrence L. Statistics for Modern Business Decisions, 4th ed. San Diego: Harcourt Brace Jovanovich, Inc., 1987. Lloyd's Register Quality Assurance, Ltd., "Getting the Most from ISO 9000." 1999. Lowenthal, Jeffrey N. Six Sigma Project Management: A Pocket Guide. Milwaukee, WI: ASQ Quality Press, 2001. Melan, Eugene H. Process Management: A Systems Approach to Total Quality. Portland, OR: Productivity Press, 1995. O'Dell, Karla, and C. Jackson Grayson, Jr. If Only We Knew What We Know. New York: Free Press, 1999. Ouelette, Steven M., and Michael V. Petrovich. "Daily Management and Six Sigma: Maximizing Your Returns." Proceedings of ASQ's 56th Annual Quality Congress, 2002. Pande, Peter S., Robert P. Neuman, and Roland R. Cavanagh. The Six Sigma Way Team Fieldbook: An Imple¬ mentation Guide for Process Improvement Teams. New York: McGraw-Hill Trade, 2001. Pyzdek, Thomas. "Six Sigma Is Primarily a Manage¬ ment Program." Quality Digest, June 1999, 26. Reilly, Norman B. The Team Based Product Develop¬ ment Guidebook. Milwaukee, WI: ASQ Quality Press, 1999. Robbins, C. L., and W. A. Robbins. "What Nurse Managers Should Know about Sampling Techniques." Nursing Management 20, no. 6 (June 1989), 46^8. Rosenfeld, Manny. "Only the Questions That Are Asked Can Be Answered," Quality Progress, April 1994, 71-73. Sherman, Strat. "Stretch Goals: The Dark Side of Asking for Miracles," Fortune, November 13,1995, 231-232.
'
they developed. In most cases, somebody just decided they were good to have. For example, IDS Financial Services, a subsidiary of American Express, used to mea¬ sure more than 4,000 individual tasks: functions like phone calls, mail coding, and application acceptance. Many of these tasks were subject to 100 percent inspection. Now, after redesigning its information management system, IDS measures 80 service processes and uses statistical sampling. Mark Graham Brown suggests some practical guidelines for designing a perfor¬ mance measurement system:16 • Fewer is better. Concentrate on measuring the vital few key variables rather than the trivial many. • Measures should be linked to the factors needed for success, namely, the key business drivers. ' y • Measures should include a mix of past, present, and future to ensure that the organization is concerned with all three perspectives. • Measures should be based around the needs of customers, shareholders, and other key stakeholders. • Measures should start at the top and flow down to all levels of employees in the organization. • Multiple indexes can be combined into a single index to give a better overall assessment of performance. Measures should be changed or at least adjusted as the environment and strategy changes. • Measures need to have targets or goals that are based on research rather than arbitrary numbers. Linking Measures to Strategy
A balanced scorecard approach helps in identifying the right measures by aligning them with the organization s vision and strategy. It provides a means of setting targets and allocating resources for short-term planning, communicating strategies, aligning departmental and personal goals to strategies, linking rewards to performance, and supplying feedback for organizational learning. IBM Rochester's quality scorecard groups 19 key performance measures into seven areas: customer satisfaction, software performance, hardware performance, service, delivery, administration, and image. Each is reported quarterly as a single index, with color-coded "quality" and "status" columns. In the quality column, red indicates an adverse trend, yellow a flat trend, and green an improving trend; in the status column, red means "not on track to attain plan," yellow means "plan attainment at risk," and green means "tracking to plan." This system provides a concise, visual summary of overall organizational performance. Effective performance measures that are aligned with business strategy are driven by factors that determine what is important to the success of the business. These factors include the following:
387
388
Part 2
• • • • • • • • •
Building Prod¬ ucts Operations
Quality in High-Performance Organizations
The nature of a company's products and services Principal customers and their key performance requirements and expectations Organizational culture; its purpose, mission, and vision Capabilities and core competencies, such as human resources, facilities, and technologies Supplier, supply chain requirements, and partnering relationships Regulatory environment Position in the market and competitive environment Principal factors that determine competitive success, such as product innova¬ tion, price leadership, or e-services Strategic challenges the organization faces
For example, the First National Bank of Chicago asked its customers what they con¬ sidered as good-quality features of a product and the delivery of those features.17 Responses included timeliness, accuracy, operations efficiency, economics, and cus¬ tomer responsiveness. These responses initiated the development of performance indicators such as lockbox processing time, bill keying accuracy, customer service inquiry resolution time, and money transfer timeliness. A computer software com¬ pany might not need to collect extensive data on environmental quality issues whereas a chemical company certainly would. A pizza franchise that delivers bulk orders to fraternities and parties around a college campus would have a different set of performance measures and indicators than one in a quiet suburban residential neighborhood. Thus, an organization first needs to fully understand its internal capa¬ bilities and external environment. Measures should logically be bed to key busi¬ The things an organization needs to ness drivers. MBNA, the Wilmington, Delaware, do well to accomplish its vision are credit card company that markets custom cards often called key business drivers to "affinity groups" such as professional associa¬ or key success factors. They rep¬ tions, universities, and sports team fans, views resent things that separate an orga¬ speed of service as one of its key business drivers. nization from its competition and Thus, it measures the time to process customer define strengths to exploit or weak¬ nesses to correct. address changes, the percentage of times phones are picked up within two rings, and the times taken to transfer calls from the switchboard.18 Armstrong Building Products Operations identified five components of value that drive its business strategy: customer satisfac¬ tion, sales growth, operating profit, asset management, and high-performance organi¬ zation. Each of these components is supported by key measurements and analysis approaches. For example, product quality, a key driver of customer satisfaction, is mea¬ sured by dimensions and squareness, fire performance, acoustics and color, dimen¬ sional stability, competitor product quality analysis, and claims. Likewise, service quality is measured by on-time delivery and missed-item promises, pricing and billing, and information support for customers. Another key business driver, operating profit, is measured by process effectiveness, units per employee, scrap and downtime, and cost of quality. Organizational performance measures include recordable injury rate, number of improvements/work orders, percentage of employees recognized, gainsharing savings, and employee satisfaction trends and turnover rate. Key performance measures should be aligned with strategies and action plans. Set¬ ting targets for each measure provides the basis for strategy deployment as discussed in Chapter 5. Figure 8.5 shows an example from Merrill Lynch Credit Corporation (MLCC), a 1997 Baldrige Award recipient. MLCC defines its Critical Few Objectives from its long-term focused strategies. Client Satisfaction, Partner Satisfaction, Business
Chapter 8
Performance Measurement and Strategic Information Management
389
O
o
05 5:
oE 13 0 Q. o
u
Q_
o o
o
CL O < CL
*
>
o 0 "c
05
o
CC
o
Q. L_ o
CJ
c
cr
o 0 0
DC
0 L_
CJ
o c
>
o 0 c 0 c
O o 0 LL
0 0 0 b —
O 6 o
00 0 i. 3 O)
LZ
o
CO T_
CD E ID
q^o
o L0_
O T•® c _Q t
LO
_Q 0 LL 0
o 0 0
TS
o
c
E
0 0 0 c 0 Q. 0 0 0 b T-
>» ^0 0 O c
0 c 0 Q. 0 O D T3 0 QC CM CM
15 0
0 0 CD 0 C 0
E CL O 0 > T5 0) O Q b
TD
Z3 Tj O Q. 0
0 T3 r 0 0 0
E
> 0
CD C k=
0 0
O
0 CL O CL 0
0 >% ZJ O 0 LL 0
uoipepijes
uojjoejsues
jueno O 'I-
j0u;jBcj O'Z
M1M0J0
an|BA
ssauisng
J0p|OLj9JBLjg
00
0>
o Q. >
C/5 .52
390
Part 2
Quality in High-Performance Organizations
Growth, and Shareholder Value. For each of these, they define key performance mea¬ sures and targets by which to evaluate progress toward meeting these objectives. They also align each of the objectives with their eight primary value-creation processes: Design, Market, Pre-Origination, Order, Uhderwrite, Approve, Audit/Fund, and Setup Service. Such an approach ensures that process owners focus on the right measure¬ ments that support the company's strategy. Process-Level Measurements
What makes a good performance measurement system? Many organizations define specific criteria for selecting measures and indicators. IBM Rochester, for example, asks the following questions: • • • • • • • • •
Does the measurement support our mission? Will the measurement be used to manage change? Is it important to our customers? Is it effective in measuring performance? Is it effective in forecasting results? Is it easy to understand/simple? Are the data easy/cost-efficient to collect? Does the measurement have validity, integrity, and timeliness? Does the measure have an owner?
Many organizations use the acronym StvlART to characterize good measures and indicators: simple, measurable, actionable, related (to customer requirements and to each other), and timely. Process measures should also clearly align with customer require¬ ments. For example, product reliability might be measured by the number of repair calls, billing Good measures and indicators are accuracy by the percentage of billing inquiries or actionable; that is, they provide complaints, knowledgeable customer represen¬ the basis for decisions at the level at tatives by supervisor observations or analysis of which they are applied. recorded calls, ease of use by number of calls to a help desk, and so on. At the process level, product and service quality indicators focus on the outcomes of manufacturing and service processes. A common indicator of manufacturing quality is the number of nonconformities per unit, or defects per unit. Because of the negative connotation of "defect" and its potential implications in liability suits, many organizations use the term nonconformance; however, quite a few still use the term defect. In this book, both terms are used interchangeably to be consistent with current literature and practice. In services, a measure of quality analogous to defects per unit is errors per opportunity. Each customer transaction provides an opportunity for many different types of errors. Nonconformities per unit or errors per opportunity are often reported as rates per thousand or million. A common measure is dpmo—defects per million opportuni¬ ties. Thus, a defect rate of 2 per 1,000 is equivalent to 2,000 dpmo. At some Motorola factories, quality is so good that they measure defects per billion! Many companies classify defects into three categories: 1. Critical defect: A critical defect is one that judgment and experience indicate will surely result in hazardous or unsafe conditions for individuals using, maintaining, or depending on the product and will prevent proper performance of the product. 2. Major defect: A major defect is one not critical but likely to result in failure or to materially reduce the usability of the unit for its intended purpose.
Chapter 8
Performance Measurement and Strategic Information Management
391
3. Minor defect: A minor defect is one not likely to materially reduce the usability of the item for its intended purpose, nor to have any bearing on the effective use or operation of the unit.19 Critical defects may lead to serious consequences or product liability suits; thus, they should be monitored and controlled for carefully. On the other hand, minor defects might not be monitored as closely, because they do not affect fitness for use. For many products, however, even minor defects can lead to customer dissatisfac¬ tion. To account for each category, many companies create a composite index in which major and critical defects are weighted more heavily than minor defects. For example, FedEx has an extensive quality measurement system that includes a com¬ posite measure, called the service quality indicator (SQI), which is a weighted sum of 10 factors that reflect customers' expectations of company performance. FedEx's SQI is shown in Table 8.1. Different weights reflect the importance of each failure; losing a package, for instance, is more serious than delivering it a few minutes late. The index is reported weekly and summarized on a monthly basis. Identifying and Selecting Process Measures
To generate useful process performance measures a systematic process is required.20 1. Identify all customers of the system and determine their requirements and expectations. Organizations need answers to key questions: Who are my customers? and What do they expect? Many of the "customer listening" approaches introduced in Chapter 4 can be used in this step. Customer expectations change over time; thus, regular feedback must be obtained. 2. Define the work process that provides the product or service. Key questions include: What do I do that affects customer needs? and What is my process? The use of
Table 8.1 FedEx Service Quality Indicator and Factors Error Type Description Weight 1. Complaints reopened— customer complaints (on traces, invoices, missed pickups, etc.) reopened after an unsatisfactory resolution 3 2. Damaged packages—packages with visible or concealed damage or spoilage due to weather or water damage, missed pickup, or late delivery 10 3. International—a composite score of performance measures of international operations 4. Invoice adjustments—customer requests for credit or refunds for real or perceived failures 5. Late pickup stops—packages that were picked up later than the stated pickup time
1
3
6. Lost packages claims for missing packages or with contents missing 7. Missed proof of delivery—invoices that lack written proof of delivery information
10
8. Right date late—delivery past promised time on the right day 9. Traces—package status and proof of delivery requests not in the COSMOS MB computer system (the FedEx "real time" tracking system) 10. Wrong day late—delivery on the wrong day
1
Source: Service Quality Indicators at FedEx (internal company document).
•]
3 5
392
Part 2
Quality in High-Performance Organizations
flowcharts for process mapping can stimulate the definition of work processes and internal customer-supplier relationships. 3. Define the value-adding activities qnd outputs that compose the process. This step— identifying each part in the system in which value is added and an intermediate output is produced—weeds out activities that do not add value to the process and contribute to waste and inefficiency. Analysis performed in this step identifies the internal customers within the process along with their needs and expectations. 4. Develop specific performance measures or indicators. Each key activity identified in step 3 represents a critical point where value is added to the output for the next (internal) customer until the final output is produced. At these checkpoints, per¬ formance can be measured. Key questions include: What factors determine how well the process is producing according to customer requirements? What devi¬ ations can occur? What sources of variability can occur? 5. Evaluate the performance measures to ensure their usefulness. Questions to consider include: Are measurements taken at critical points where value-adding activi¬ ties occur? Are measurements controllable? Is it feasible to obtain the data needed for each measure? Have operational definitions for each measurement been established? Operational definitions are precise definitions of measure¬ ments that have no ambiguities. For example, when measuring "invoice errors," a precise definition of what is an error and what is not is needed. Does an error include an omission of information, wrong information, or misspelling? Operational definitions provide a common understanding and enhance com¬ munication throughout the organization. To illustrate this approach, consider the process of placing and filling a pizza order. Customer expectations include a quick response and a fair price. The process that pro¬ vides this service is shown in Figure 8.6. To begin, the order taker is an (internal) cus¬ tomer of the caller (who provides the pizza order). Later, the caller is a customer of the deliverer (either at the pickup window or the caller's home). Also, the cook is a cus¬ tomer of the order taker (who prepares the documentation for the ordered pizza). Some possible performance measures include: • Number of pizzas, by type per hour. If this number is high relative to the kitchen's capacity, then perhaps cooking time and/or preparation is being short-cut or delivery times are stretched out. • Order accuracy (as transmitted to the kitchen). This measure can indicate a lack of attention or knowledge on the part of the order taker. • Number of pizzas rejected per number prepared. A high number for this mea¬ sure can indicate a lack of proper training of cooks, resulting in poor products and customer complaints. • Time to delivery. This measure might indicate a problem within the restaurant or inadequate training of the driver. (Of course, as happened with Domino's, measuring delivery time could encourage drivers to drive too fast and lead to safety problems.) • Number of errors in collections. Errors here can result in lost profits and higher prices. • Raw materials (dough, etc.) or finished pizzas inventory. A high number might result in spoilage and excess costs. Low inventory might result in lost orders or excessive customer waiting time. Notice that these measures—only a few among many possible measures—are related to customer expectations and business performance.
Chapter 8
Performance Measurement and Strategic Information Management
Figure 8.6 Example of a Pizza Ordering and Filling Process for Home Delivery
Many organizations use dashboards, which typically consist of a small set of measures (five or six) that provide a quick summary of process performance (this term is sometimes also used to describe a balanced scorecard at the organizational level). This preference stems from the analogy to an automobile's dashboard—a col¬ lection of indicators (speed, RPM, oil pressure, temperature, etc.) that summarize performance. Dashboards often use graphs, charts, and other visual aids to commu¬ nicate key measures and alert managers when performance is not where it should be. Aligning Strategic and Process-Level Measurements21
It is possible that all work processes could be meeting their requirements while the organization is not achieving its longer-term goals. Thus, aligning strategic and process-level measurements is vital to a high-performing organization, and can be viewed as an approach for strategy deployment (see Chapter 5). Figure 8.7 illustrates how goals and measures might be aligned for a hypothetical retail manufacturer. Alignment might even go further, down to the team and individual levels. Note that alignment is tied fundamentally to the performance goals; the measures support goal attainment. The organization does not have to have one set of performance measures that everyone produces and reports, but rather, measures are used where they are
393
394
Part 2
Quality in High-Performance Organizations
Figure 8.7 An Example of Aligning Strategic and Process-Level Performance Measures Corporate Level
Source: R. I. Wise, "A Method for Aligning Process-Level and Strategy-Level Performance Metrics," The Quality Management Forum, 25, no. 1 (Spring 1999), 4-6. American Society for Quality, 11th Annual Quality Management Conference.
most appropriate. Production line data, for example, might be reviewed only at the line level for daily operations control, while some data might be integrated at the next level for process improvement. Information that supports review of organizationallevel performance is passed on to the corporate level. Enterprise Resource Planning (ERP)—systems are software packages that inte¬ grate organizational information systems and provide an infrastructure for man¬ aging information across the enterprise.22 They integrate all aspects of a business—accounting, customer relationship management, supply chain manage¬ ment, manufacturing, sales, human resources, and so on—into a unified information system, and provide more timely analysis and reporting of sales, customer, inven¬ tory, manufacturing, human resource, and accounting data. The three most promi¬ nent vendors for ERP software are SAP, Oracle, and PeopleSoft. For example, when a salesperson fills an order, the system can check the customer's credit and their own manufacturing capacity, record the order, schedule the shipment, log the order on the production schedule, order parts from suppliers, and update financial and accounting records. ERP systems allow companies to share different databases in a networking environment and store and process all company data in a unique data¬ base, and distribute it to a large group of users. Typical ERP applications span finan-
Chapter 8
Performance Measurement and Strategic Information Management
cial, human resource, operations and supply chain, and sales and marketing data. Many ERP systems now offer performance measurement system modules that are focused on helping manage the wide scope of data that are collected in the system.
ANALYZING AND USING PERFORMANCE DATA
All the different types of data that we discussed support operational-level decisions, semor leadership performance reviews, priority setting, and strategic planning! However, simply reporting numbers or showing them on graphs and charts are not enough. Data require sound analysis to turn them into information. Analysis refers to an examination of facts and data to provide a basis for effective decisions. Exam¬ ples of possible analyses include the following: • Examining trends and changes in key performance indicators • Making comparisons relative to other business units, competitor performance, or best-in-class benchmarks Calculating means, standard deviations, and other statistical measures Seeking to understand relationships among different performance indicators using sophisticated statistical tools such as correlation and regression analysis The capabilities of today's spreadsheet and database software, such as Microsoft Excel and Access, make analysis simple to do by nearly any employee. Also, some evidence suggests that organizations that use more sophisticated statistical tools for analysis tend to have better business results. The fact that effective analysis requires more advanced statistical thinking might explain the lack of good approaches in most organizations. Thus, organizations are advised to develop improved statistical exper¬ tise among their employees which is one of the key benefits of Six Sigma. Volumes of data acquired at the process level, while useful for daily operations decisions and process control, generally are not appropriate for senior executive review or strategic planning. For instance, some companies develop an aggregate customer satisfaction index (CSI) by weighting satisfaction results, market share, and gains or losses of customers. Previously we discussed how FedEx aggregates dif¬ ferent quality components into a single index. Corning Telecommunications Prod¬ ucts Division aggregates data and information into key financial and business-level analyses that quantify the impact of decisions on financial and market performance. Its data researchers examine relationships among quality, price, and image as they relate to the attraction and retention of customers. They also quantify the relation¬ Organizations need a process for transforming data, usually in some integrated fashion, into information that top management can under¬ stand and work with.
ships between process capability, people and process productivity, and unit costs. As we noted in our discussion of the bal¬ anced scorecard, managers must also under¬ stand the linkages between key measures of business performance. Examples of such analyses are:
• How product and service quality improvement correlates with key customer indicators such as customer satisfaction, customer retention, and market share • Financial benefits derived from improvements in employee safety, absenteeism, and turnover • Benefits and costs associated with education and training • Relationships between product and service quality, operational performance indicators, and overall financial performance
395
396
Part 2
Quality in High-Performance Organizations
• Profit impacts of customer satisfaction and retention • Market share changes as a result of changes in customer satisfaction • Impacts of employee satisfaction on customer satisfaction
and Light
Establishing causal relationships between external lagging results and internal leading measures provides a visible and obvious direction for improvement. For instance, FedEx correlates the 10 quality components with customer satisfaction through extensive market research. Additionally, it conducts focus groups and other surveys to validate these relationships. Trident Precision Manufacturing correlates quality, cost, delivery, and service data with customer satisfaction data to verify their understanding of their customers. GTE Directories uses a proprietary third-party model and software to perform similar analyses. Interlinking is the term that describes the quantitative modeling of cause-andeffect relationships between external and internal performance measures, such as the relationship of customer satisfaction measures to internal process measures (e.g., product quality or employee performance).23 A simple interlinking model was devel¬ oped by Florida Power and Light.24 In studying the telephone operation of its cus¬ tomer service centers, FPL sampled customers to determine their level of satisfaction with waiting times on the telephone. Satisfaction began to fall significantly at about two minutes (see Figure 8.8). FPL also found that customer satisfaction is directly related to how callers perceive the competence of the phone representatives. Research showed that excessive waiting times caused a bias in the ratings. Eliminating this bias made a substantial contribution toward accurately measuring customer satisfaction with the phone contact experience. To improve customer satisfaction, FPL developed a system to notify customers of the anticipated wait and give them a choice of holding
Figure 8.8 Interlinking Model of Customer Satisfaction and Time on Hold
Customer Satisfaction Rating
Time on Hold
1.32
5ml/hr dis¬ crepancy). Ten of the audits were rated at level 2 and four were rated at level 3. Root cause analysis was employed to determine the cause of the dis¬ crepancies. Work was then begun to affect the accuracy of infusion rates. Using Six Sigma methods and statistical tools, the team also examined the hospital's clinical labo¬ ratory process. Key elements in the acquisition, laboratory analysis, and reporting of patient speci¬ mens were identified. The steps included (1) physician order, (2) order entry, (3) matching the order to the patient, (4) collecting the specimen, (5) labeling the specimen, (6) transporting the spec¬ imen, (7) analyzing the specimen, (8) reporting the results, and (9) entering the results into the patient's chart. Each of these steps is subject to error. Applying Six Sigma analysis, the steps sub¬ ject to the most errors were identified. These steps
501 were: order entry by the unit clerical staff, trans¬ portation of the specimens to the lab, and analysis of specimens in the lab. To identify, define, and reduce these errors, a laboratory error reduction task force was established. It included members from administration, lab, nursing, clerical staff, information systems, and quality management. The task force first developed a process map so that all members could appreciate the complexity and vulnerability of the entire process. The process map provided the task force with the tools to ana¬ lyze the clinical laboratory problem in depth. The FMEA technique was employed to arrive at a risk priority number (RPN) so that steps in the labora¬ tory analysis process could be prioritized in terms of their vulnerability to error. Again, order entry, transportation, and analysis of specimens were identified. Statistical tools, including correlation and regression, analysis of variance, confidence intervals, and hypothesis testing, were employed to evaluate the laboratory process further. The analysis of medication delivery by IV infusions served as a good example of deployment of Six Sigma methodology to reduce error and improve patient safety in a health care setting. Sig¬ nificant variability in the ordering and processing of IV drips was identified. Lack of standardization in many steps of the process posed the greatest risk for system failure. Those steps with the highest degree of variability and the greatest chance for error were 1. MD ordering practices (i.e., lack of standard¬ ization in medication description, dosage, concentration, etc.) 2. IV drip preparation (lack of standardization by pharmacy and nursing of IV bag concen¬ trations) 3. RN labeling and documentation of IV con¬ centrations In these three areas, a multidisciplinary task force created standards to reduce variation. Spe¬ cific interventions included implementation of standardized physician order sheets, a policy requiring preparation of all IV medications in a standard concentration, and use of color-coded labels when nonstandard concentrations were in use. Thirty days after implementation, measurable improvement was evident. Level 1 discrepancies fell from 47.4 percent to 14 percent. Level 2 dis¬ crepancies fell from 21.1 percent to 11.8 percent
502
Part 3
and level 3 discrepancies fell from 15.8 percent to 2.9 percent. Though far from achieving a six-sigma level of performance, substantial efforts continue to move toward that goal. The laboratory project proved to be more com¬ plex. It was evident early on that the scope of this complex system was too broad for an initial effort. The project was broken down into smaller indi¬ vidual steps of the larger process. Once refocused, the appointed task force identified opportunities to reduce variation in select steps of the laboratory process. Alternate means of identifying specimens, changes in the approach to "point of care" labora¬ tory analysis, decentralization of some laboratory tests, and a revised system to order and process
Six Sigma and the Technical System
stat lab tests was put into place. Effectiveness monitoring continues as does measurement of sus¬ tainable error reductions. These efforts marked the beginning of a long laboratory redesign process aimed at driving out error, reducing turnaround time, and improving patient safety.
Key Issues for Discussion 1. How did the team use process mapping as a key part of the Six Sigma process? What value did process mapping have? 2. Why were the teams and task forces multi¬ disciplinary in nature? What benefits does this approach have?
Quality in Practice Ford’s Drive to Six Sigma Quality36 Ford Motor Company began developing its Six Sigma quality approach, called Consumer Driven Six Sigma, in 1999. However, the company didn't really get serious about reclaiming their motto of the 1980s, "Quality is Job 1," until 2001—when JD Power and Associates' Initial Quality Study ranked Ford last among the big seven automakers. By 2003, the same survey ranked Ford number four and found that they were the most improved auto maker of the group. The company now has more than 200 Master Black Belts, 2200 Black Belts, nearly 40,000 Green Belts, and 3000 Project Champions. Ford's training of Green, Black, Master Black Belts, and Project Champions generally follows the conventional Six Sigma training process. Black Belt training is "hands-on" and "just in time." Each trainee gets one week of full-time training per month for four months. The other three weeks of the month require that the trainees apply their training to a live project. Their Six Sigma teams typically have a member of management, a Master Black Belt (MBB), a Black Belt (BB), and several Green Belts (GB) assigned to take on various roles in a project. BBs are expected to handle two to three projects at a time. They can choose their own projects, but are asked to choose them carefully to ensure that they contribute to waste elimination or customer satisfaction improvement. The goal is that at least
half of the reduction of "Things Gone Wrong" (in "Ford-speak") will be improved through successful Six Sigma projects. Ford has implemented a unique project tracking system that has helped to promote organizational learning. The system allows mem¬ bers of project teams to observe what other teams are working on via an internal database. Leaders are also expected to have hands-on involvement as project champions. Senior leaders are required to partner with MBBs to run perfor¬ mance cells. These cells are managed similar to a manufacturing operation and benefit from the technical expertise of the MBB and the administra¬ tive experience of the manager. The process keeps new projects coming in and ensures that projects that are underway stay on track. An example of a typical project was the one led by Master Black Belt Pauline Burke. The problem was recognized after customers complained that body side moldings on the Ford Focus were lifting at the edges. It became evident to Burke that this issue was a "mega project" when the number of CTQ (critical to quality) issues began to multiply. In total, the project required nine months to com¬ plete, compared to the average of four months for a typical Six Sigma project at Ford. Burke and her team followed the DMAIC problem-solving process rigorously. The Define stage uncovered four critical issues:
Chapter 10
Principles of Six Sigma
1. The tape that was designed to secure the molding was not contacting the car body enough. 2. Holes located on the body and used to line up the molding were too high, hitting an indent on the body sides. 3. Pressure used to apply the tape was too low. 4. The body was not clean enough, so the tape was not sticking well. hi the Measure stage, measurements were taken on the location of the holes, flatness of the molding, pressure being applied, and percent of area being cleaned. Analysis required that team experts, stake¬ holders such as maintenance personnel and tier 1 and tier 2 suppliers, as well as management, use the data. They were all seeking to understand the process and to discover ways in which it could be improved. In Stage 4 of DMAIC, improvements were proposed, including moving holes on the body side down by 2 millimeters; changing molds for the body side molding to ensure flatness and 100 percent contact between the molding, tape, and body side; using optimum pressure to apply the molding (as determined by a design of experiment process); and replacing the head on the cleaning fix¬ ture to ensure optimum cleaning of the body side. One element of the Control stage was to mon¬ itor the hole locations using routine quality checks. It was also necessary to ensure that the supplier implemented a new procedure for checking the moldings for flatness. Other quality checks were performed to meet specifications for optimum pres¬ sure used to apply the moldings to the body, and to maintain cleaning equipment. The project resulted in savings of $100,000 and no customer complaints since the improvements were implemented. Overall, Ford's Six Sigma approach contributed impressively to the bottom line. More than 6,000 projects have been completed in just three years, and Six Sigma has saved more than $1 billion since its inception. Louise Goeser, Ford's vice president of
jjgfl
503 quality, cited a number of benefits of the company's Six Sigma effort, including improved quality of products, better measurement of results and suc¬ cess, and improved decision making. However, she also noted some challenges, such as selecting pro¬ jects that are linked to strategic objectives and the company's Revitalization plan. Ford's goal is corpo¬ ratewide adoption of Six Sigma tools and method¬ ology, so that everyone from the CEO down will possess data-driven decision-making skills. The slogan, "Quality is Job 1," has been given a new emphasis with the development of three components: operating systems to define standards and processes, quality leadership to engage all employees, and Consumer Driven Six Sigma to be the primary data-driven decision process. These ele¬ ments have helped to integrate Six Sigma into the overall quality program. Finally, the company plans to continue its emphasis on value creation and waste prevention, while widening and deepening deployment. This focus will involve increasing use of Design for Six Sigma, strengthening ties with sup¬ pliers, and continued integration of Six Sigma tools, methods and mindset as a mechanism for delivering results based on corporate objectives. Key Issues for Discussion
1. Why do you think that the 2001 J.D. Power and Associates results were so poor even though the company had started its Six Sigma process in 1999? 2. Why did it take almost twice the average project time in order to complete the Ford Focus body-molding project? What were some possible technical difficulties encoun¬ tered in analyzing and correcting the CTQ issues that were uncovered in the design stage of that project? 3. A major roadblock in Ford's Six Sigma effort was employee skepticism. How do you think they overcame it?
Review Questions
1. What is a defect? Explain how to compute defects per million opportunities (dpmo). 2. Explain the theoretical basis for Six Sigma quality. How does it relate to the process capability index Cp?
504
Part 3
Six Sigma and the Technical System
3. Describe the Six Sigma problem-solving approach (DMAIC). How is it similar to or different from the other problem-solving approaches discussed in this chapter? 4. What are the key principles for effective implementation of Six Sigma? 5. What are the major types of tools used in Six Sigma projects? 6. What is Kepner and Tregoe's definition of a problem? How does this definition apply to quality issues? Provide some examples. 7. Explain the difference between structured, semistructured, and ill-structured problems. What implications do these classifications have for solving prob¬ lems? 8. What are the four major components of problem solving? Why is it important to have some type of systematic problem-solving methodology in an organiza¬ tion? 9. List and explain the five categories into which all quality problem-solving can be classified. 10. Why do messes arise in organizations? 11. What is a root cause? How does the "5 Why" technique help uncover the root cause? 12. Describe some techniques used to generate ideas. 13. List and explain some of the tools and approaches used in "lean" organizations. How does the lean operating concept relate to Six Sigma? 14. What are some reasons why the lean approach appeals to small organiza¬ tions? 15. In manufachrring, the concept of a "hidden factory" describes the necessity of repair and rework of defective products. List some places where the "hidden factory" can exist in service businesses.
1. The January 22,2001, issue of Fortune contained an article "Why You Can Safely Ignore Six Sigma," that was highly critical of Six Sigma. Here are some of the criticisms levied against Six Sigma: a. The results often don't have any noticeable impact on company financial state¬ ments. Thus, Six Sigma success doesn't correlate to higher stock value. This criticism applies to 90 percent of the companies that implement Six Sigma. b. Only early adopters can benefit. c. Six Sigma focuses on defects, which are hard to objectively determine for ser¬ vice businesses. d. Six Sigma can't guarantee that your product will have a market. How would you respond to these statements? 2. Some of the key processes associated with business activities for a typical com¬ pany include sales and marketing, supply chain management, managing infor¬ mation technology, and managing human resources. What types of Six Sigma projects might be considered in order to improve each of these activities? 3. "Resistance to change" is a common theme in the behavioral sciences. What part do you believe that resistance to change plays in management's fostering of successful versus unsuccessful adoptions of Six Sigma approaches? What impact does workers' resistance or lack of resistance have?
Chapter 10
Principles of Six Sigma
4. List some of the common processes that a student performs. How can these processes be improved using a Six Sigma approach? 5. Why are modern products that often require high tolerances, short production runs, and heavy customer input difficult to manufacture in order to meet Six Sigma specifications? 6. How can lean concepts be applied in a classroom? 7. The Six Sigma philosophy seeks to develop technical leadership through "Belt" training, then use it in team-based projects designed to improve processes. To what extent are these two concepts (technical experts versus team experts) at odds? What must be done to prevent them from blocking success in improve¬ ment projects? 8. How might a Six Sigma project be done to improve a registration process in a university? An admission process? 9. How can a manager effectively balance the key components of a Six Sigma implementation design related to who, what, where, when, why, and how it could be done? 10. In 1995 Jack Welch sent a memo to his senior managers telling them that they would have to require every employee to have started Six Sigma training to be promoted. Furthermore, 40 percent of the managers' bonuses were to be tied to the successful introduction of Six Sigma. Do you believe that this directive was a motivational action, or did it violate W. Edwards Deming's maxim that man¬ agers and leaders must "cast out fear"? Why or why not? 11. A consultant told the story of two Six Sigma teams that made separate presenta¬ tions on how they would improve processes in their own areas. At the end of the second presentation, the consultant asked a basic question that stopped both Black Belt team leaders in their tracks: "Haven't you both just proposed making improvements based on eliminating parts of processes in the other group's areas? It seems that the implementation costs in one area will cancel out the savings in the other area!" What had the Black Belts failed to recognize? What would you recommend to prevent this situation from happening in other organizations?
Problems 1. An insurance firm set a standard that policy applications be processed within three days of receipt. If, out of a sample of 1,000 applications, 50 fail to meet this requirement, at what sigma level is this process operating? 2. During one month, 35 preflight inspections were performed on a military air¬ craft. Eighteen nonconformances were noted. Each inspection checks 60 items. What sigma level does this incidence of nonconformance correspond to? 3. Over the last year 1,054 injections were administered at a clinic. Quality is mea¬ sured by the proper amount of dosage as well as the correct drug. In two instances, the incorrect amount was given, and in one case, the wrong drug was given. At what sigma level is this process? 4. The Wall Street Journal reported on February 15, 2000, that about 750,000 air¬ plane components are manufactured, machined, or assembled for Boeing Co. by workers from the Seattle Lighthouse for the Blind. A Boeing spokeswoman noted that the parts have an "exceptionally low" rejection rate of one per thou¬ sand. At what sigma level is this process operating?
505
506
Part 3
Six Sigma and the Technical System
5. An electronics firm manufactures 500,000 circuit boards per month. A random sample of 5,000 boards is inspected every week for five characteristics. During a recent week, two defects were found for one characteristic, and one defect each was found for the other four characteristics. If these inspections produced defect counts that were representative of the population, what is the overall sigma level for the process? What is the sigma level for the characteristic that showed two defects?
^
Projects, Etc 1. Three popular Web sites for Six Sigma are http://www.ge.com/sixsigma, http://www.isixsigma.com, and http://www.sixsigmaformum.com. Explore these sites and consider the following questions. a. How does GE use Six Sigma to enhance customer perception of its products and services? b. What is the apparent purpose of the isixsigma Web site? c. Who are the customers of the sixsigmaforum Web site? d. Is there basic agreement about what the Six Sigma concept includes, based on what the three Web sites present? Why or why not? e. How do these Web sites differ in their concept of what Six Sigma includes? 2. Identify an important problem around your school or in some related function, such as a student organization, and apply the DMAIC process to develop an improved solution. You might wish to consult Chapter 13 during this effort to learn some useful tools. 3. Find a local company that is using Six Sigma or lean principles. Write a case study of their experiences, focusing on the challenges they faced during their implementation efforts.
^ Cases I. Implementing Six Sigma at GE Fanuc37 GE Fanuc Automation in Charlottesville, Virginia, is a joint venture between General Electric and Fanuc Ltd. of Japan, a company that specializes in com¬ puter numerical control (CNC) and robotic tech¬ nology. The division has annual sales of about $700 million from the manufacture and sale of factory automation products, which serve the automotive, food-processing and packaging, paper, pharmaceu¬ tical, robotics, chemical, and energy markets. The headquarters and main manufacturing plant is at its Charlottesville facility, and includes more than 500,000 square feet of floor space divided among seven buildings on 50 acres of land. GE Fanuc implemented their Six Sigma program in 1996, shortly after Jack Welch announced the quality ini¬
tiative for the entire company. The program required a major cultural and attitude change at GEFanuc and around the world at GE sites, but it has resulted in a stronger, quality-driven company. The Six Sigma way of thinking is ingrained in everything the company and its employees do. "From our corporate decisions all the way out to the factory floor. Six Sigma has raised our employees' mindset to look at data instead of emotion," says Sheila O'Donnell-Good, GE Fanuc's Six Sigma busi¬ ness leader. "If you go out on the floor and visit each line, you're going to see a lot of good data driving decision making.. .. We have ingrained our tool sets within our people, so Six Sigma is a philosophy and an outlook that allows us to examine a broken
Chapter 10
Principles of Six Sigma
process, get to a solution, and put controls on in the end. We also see it as a business strategy that helps us gain a competitive edge because it's a differen¬ tiator between us and our competitors." "At one time, GE was a Three-Sigma company and the cost of failure was estimated at 15 percent of sales. But achieving Six Sigma represents a $4 billion cost reduction opportunity through reduced cost of failure," says O'Donnell-Good. She adds that the savings are "really greater if you think about it because there have been significant improvements through this program other than the cost-of-failure reduction." Six Sigma teams are established to improve or correct processes. Don Splaun, manager of advanced manufacturing technology, headed a Six Sigma team that wanted to eliminate the Environmental Stress Screen (ESS) test on circuit boards. Splaun felt the test was costly and unnecessary because the ESS was followed by a second and final test. The test was designed to eliminate premature failure in the boards, but required running the boards through a high-temperature oven for seven hours. Initially, Splaun estimated that GE Fanuc was paying about $12,000 to $18,000 in electricity plus $2,000 to $70,000 a year in maintenance costs per oven and labor costs for loading and unloading the oven. Concentrating on the field-control product line, team members collected and ana¬ lyzed data to determine whether the final test was as effective as the ESS. Operators filled out data sheets with information such as board name, date, and whether the board passed or failed the ESS test and subsequent tests. These data helped team members determine whether boards that failed were false failures or dead on arrivals (DOAs), which aren't related to the ESS. Of 7,703 boards that were tested, 311 failed in the first pass. Of these, 284 (91.3 percent) were false failures and 26 (8.4 percent) were dead on arrival (DOA). Only 1 board (0.3 percent) actually failed during the ESS. DOAs were also found bad at the final test, indi¬ cating that the final test is an effective screen. Thus, Splaun and his team found only 1 failure out of 7,703 units, which was equivalent to 130 defects per million observations (dpmo), a yield of 99.99 percent, and a sigma level of 5.15. This analysis indicated that the final test cap¬ tured the same failures as the ESS in a more timeand cost-effective manner, so the ESS and the ovens used for the test could be eliminated. To con¬
507 trol the improvement, the company began to track the number of failures and defective boards on the line to ensure that product quality remains high after elimination of ESS. The actual benefits that resulted from the project are summarized here.38 Direct Labor & Materials Savings $84,742 Inventory Reduction $48,400 Energy/Maintenance $16,000 Total Hard Savings $149,142 Labor Cost Avoidance $18.000 Total Savings $167,142 Removing the test from the manufacturing process also reduced the cycle time by a day. GE Fanuc is only one example of the application of Six Sigma within General Electric. The impact of Six Sigma across the GE corporation is clearly described in the company's 1999 Annual Report:
In 1999, the Six Sigma initiative was in its fifth year—its fifth trip through the oper¬ ating system. From a standing start in 1996, with no financial benefit to the Company, it flourished to the point where it produced more than $2 billion in bene¬ fits in 1999.39 Jack Welch, then CEO of GE stated: "We want being a product/services customer of GE to be analogous to bringing your car in for a 50,000-mile check and driving out with 100 more horsepower, better gas mileage and lower emissions." In the initial stages of Six Sigma, the com¬ pany's effort consisted of training more than 100,000 people in its science and methodology and focusing thousands of "projects" on improving efficiency and reducing variance in internal operations—from industrial factories to financial services back rooms. From there, the firm's operating system steered the initiative into design engineering to prepare future generations of "Design for Six Sigma" products—and drove it rapidly across the customer-interactive processes of the financial services businesses. Medical Sys¬ tems used it to open up a commanding technology lead in several diagnostic platforms and achieve dramatic sales increases and customer satisfaction improvements. Every GE product business and financial service activity [now] uses Six Sigma in its product design and fulfillment processes. Welch concluded: "Today, Six Sigma is focused squarely where it must be—on helping our cus¬ tomers win. A growing proportion of Six Sigma
508
Part 3
projects now under way are done on customer processes, many on customer premises. The objec¬ tive is not to deliver flawless products and services that we think the customer wants when we promise them—but rather what customers really want when they want them."
Six Sigma and the Technical System
2. What is the difference between direct labor savings and labor cost avoidance savings from a managerial perspective? 3. Verify that the number of defective boards found in the test (one) gives a dpmo of 130. 4. If you were Splaun and were asked to make a presentation to other team leaders and man¬ agers (which, in fact, happened), what con¬ clusions would you draw that may be useful to future teams about the way that the pro¬ ject was conducted?
Discussion Questions
1. How was GE's corporate-level vision of Six Sigma put into practice at the GE Fanuc man¬ ufacturing site?
II. The PIVOT Initiative at Midwest Bank,40 Part 2
This case is a continuation of PIVOT Initiative at Midwest Bank in Chapter 7. You should review that case for background on the project and a descrip¬ tion of the Define stage of the DMAIC approach. In this case, the focus is on the remaining steps—Mea¬ sure, Analyze, Improve, and Control. The Measure stage demanded an intense data collection effort by the PIVOT team. They used a tool called an XY matrix (see example in Table 10.2), designed to rank factors for potential error causes (Xs) and for customer outputs (Ys). The team gath¬ ered data and studied departmental process flows, seeking to find root causes of the problem, and to identify and agree on key CTQs that impact the customer. During the process, it was difficult for everyone, especially the subject matter experts (SMEs), to ignore their perceptions in speculating about possible causes of errors, which became known as "tribal knowledge." Six Sigma theory strongly discourages any attempt to let unproven assumptions creep into recommendations. All fac¬ tors must be statistically proven through in-depth analysis to justify recommendations. For CPD, the principal customer outputs selected were risk mitigation, error reduction, and reducing dollar loss, which were then stratified against potential error causes. The matrix was then used to calculate an overall ranking to guide the team toward the most probable causes of errors. After deciding to focus on the top seven potential causes, the CPD department's staff began the task of gathering data for each of them to verify their impact on errors in the process. The Six Sigma analyst stratified the data col¬ lected across the potential error categories. During the Analysis stage, analysis of an extensive array of
graphs developed from the data permitted the team to see trends in the process and to begin seeking strategic solutions. Construction of the graphs required more than 48 hours of team effort. Trends pointed to problems with the manual strapping process. However, the tribal knowledge suggested that errors were due to insufficient staffing, but the initial analysis of the data did not match this hypothesis. As a result, the team began to search for a way to prove or disprove the tribal knowledge. It was suggested that the data were not prop¬ erly stratified with regard to staffing and that analysis should be applied across a longer timeframe. CPD set out to collect more data from past months, and the Six Sigma analyst began devel¬ oping the graphs needed to examine the new data. Over the next week, some 100 different graphs were created, depicting data in single strands and also paired with variables that interacted with one another. The team's Six Sigma training had emphasized the importance of fully exploring all data interactions, using graphs to illustrate rela¬ tionships between variables. Despite the team's best efforts to find a relationship, staffing and volume did not appear to affect strapping errors. This finding disproved the tribal knowledge, while providing a multitude of additional graphs for analysis. Strong correlations were seen in the graphs involving human factors and manual processes. CPD's processes called for numerous manual steps when handling cash. The graphs' trends suggested that whenever a manual process occurred, the number of errors increased, especially in the cash strapping area where, despite many years of experi¬ ence, associates were making more than 100 errors
Chapter 10
Principles of Six Sigma
509
Table 10.2 Examples of an XY Matrix
XY Matrix
10
% Rank
Reduce Defects
10
Rank
Mitigate Risk
Potential
Reduce Loss
Output
Variables (Ys)
Project: CDP Pivot_
Customer Compliance
9
10
10
281
15.11%
Experience
10
10
9
280
15.05%
Manual Processes
10
9
10
280
15.05%
Human Factor
10
8
10
280
15.05%
Training
8
10
10
272
14.62%
Volume
9
9
10
271
14.57°h
Interdepartmental Processing Flow
10
10
5
240
12.90%
Timeliness of Courier
3
5
8
157
8.44%
Timeliness Standards
3
3
9
147
7.90%
Staffing
5
3
6
135
7.26%
Theft
2
4
2
78
4.19%
Output Rating
Input Variables (Xs)
9
Association Table
each year. These errors caused the bank to lose thou¬ sands of dollars through miss-strapped cash. On the deposit side, manual errors created a much larger dollar loss per error. The one anomaly loss accounted for almost $280,000 dollars without any repercussions for the associate making the error. A manual process caused the error, but several team members felt that the overall attitude toward dollar errors was insufficient. Associates on the deposit side were far more concerned about quantity of deposit errors, than they were about the dollar losses from each error. These human factor elements began to cause the team great concern, because it was likely to be difficult to come to agreement on immediate solutions for such a complicated issue. To further evaluate the process, the team decided to utilize an advanced Six Sigma tool called the Failure Modes and Effects Analysis
(FMEA, discussed further in Chapter 12). Tire FMEA paralleled the process map constructed in the Measure stage, but concentrated more on the inputs to the processes. Once the steps were laid out, the team brainstormed potential fallouts or errors from the process. Each of these errors was then charted until the potential effect of the indi¬ vidual problem was found. After the causes and effects were mapped out, each process step was then ranked on three categories: severity, occur¬ rence, and detection, to create an overall ranking of potential failures (RPN) in the process and spearhead the team's efforts (see the example in Table 10.3). The highest-ranking index of the team was strapping cash, with an RPN of 360. Of the top 10 potential errors, 77 percent involved human factors as the root cause of the problem. These issues focused the team on the need to alleviate
510
Part 3
Six Sigma and the Technical System
Table 10.3 Typical Pivot FMEA Showing Key Process Steps Potential
Potential
%
Failure
Potential
Causes
Process
Modes
Failure
of
Current
Recom¬
Person and
Function
(process
Effects
Failures
Process
mended
Target
Taken
#
(step)
(defects)
(Ys)
SEV
(Xs)
occ
Controls
Actions
Date
Actions
2
Customer Makes Deposit
No deposit ticket
Deposited in wrong account
10
Human factor
3
Magnet verifies A/C# and name
14
24
31
33
51
Check Deposits Customer to fraud Processing
Bank takes loss
9
Processor Verifies Deposit
Bank takes loss
9
Bank takes loss
9
Bank takes loss
9
Lost check
Cash In Ticket/Ship Processing
Lost deposit
Strap Cash
Lost cash
Processing Completes Deposit
Miss post
Deposit delayed or bank loss
8
Bank takes loss
Human factor
Human factor
Human factor
Human factor
Responsible
2
Currently verify payee/ systemic controls
DET RPN
8
Extensive 180 customer ed. possible fee assessments
10
180
Dollar 180 standard and dual control
2
Research and correct if possible
10
1
Research and correct if possible
1
9
10
360
8
256
4
4
the human interaction with the process, and espe¬ cially on fixing strapping errors. With statistically proven error causes available, the PIVOT team turned its attention to the Improve stage to develop corrective actions. One of the most beneficial tools used to find solutions to error causes was the Countermeasures Matrix. A portion of this matrix is shown in the form of the Countermeasures Tree Diagram in Figure 10.4. This diagram helped the team organize potential solutions to the most risky issues and to ensure that root causes were effectively addressed. The diagram categorized the proposed solutions by effectiveness and feasibility on a scale of 1 to 5, based on team opinion, statistical information, and cost estimates. Some of the solu¬ tions emerged in earlier stages of the process, while others came after intense scrutiny.
Manual process
Manual process
Data collection concerning all deposited checks
SMEs Sept.
SMEs
Verifying deposit ownership and check verify °° the distribution of the random variable z = (x - |x)/(a/v n) approaches that of a standard normal distribution.
The power of the central limit theorem can be seen through computer simulation using the Quality Gainebox software that is included on the CD-ROM accompanying
Chapter 11
Statistical Thinking and Applications
539
this text. Figure 11.14 shows the results of sampling from a triangular distribution for sample sizes of 1, 2, 5, and 10. For samples as small as five, the sampling distribution begins to develop into the symmetric bell-shaped form of a normal distribution. Also observe that the variance decreases as the sample size increases. The approximation to a normal distribution can be assumed for sample sizes of 30 or more. If the population is known to be normal, the sampling distribution of x is normal for any sample size. Next, consider the sampling distribution of p, in which the expected value of p, E{p) = n. Here n is used as the population parameter and is not related to the number n ~ 3.14159. The standard deviation of p is
sV
7t(l — 7T)
for infinite populations. For finite populations, or when n/N > 0.05, modify sp by sV
N-n
n(l-n)
N-l
In applying the central limit theorem (CLT) to p, the sampling distribution of p can be approximated by a normal distribution for large sample sizes. This and subsequent chapters explore various applications of the CLT to statis¬ tical quality control in the areas of process capability determination and control charting. Consider the following example as we illustrate an application of sampling distributions. The mean length of shafts produced on a lathe has historically been 50 inches, with a standard deviation of 0.12 inch. If a sample of 36 shafts is taken, what is the probability that the sample mean would be greater than 50.04 inches?
Figure 11.14 Illustration of the Central Limit Theorem
Theoretical Distribution
Actual Distribution
Actual Distribution
Sample size = 1
Sample size = 2
Actual Distribution
Actual Distribution
Sample size = 5
Sample size = 10
Source: Courtesy of P-Q Systems, Inc.
540
Part 3
Six Sigma and the Technical System
The sampling distribution of the mean is approximately normal with mean 50 and standard deviation of 0.12/V36. Thus, =
X-JL
=
o/VS
50,04-50 =
4
0.12/V36
In the standard normal table, the value of 2.0 yields the probability of 0.4772 between the mean and this value. The area for z > 2.0 then is found by P(z > 2.0) = 0.5000 - 0.4772 = 0.0228 Thus, the probability of a value equal to or greater than 50.04 inches as the mean of a sample of 36 items is only 0.0228 if the population mean is 50 inches. The applica¬ bility of sampling distributions to statistical quality is that "shifts" in the population mean can quickly be detected using small representative samples to monitor the process. Similarly, if a sample size of 64 is used, a/-y/n = 0.12/8 = 0.015 and Z
=
x - il —£ g/V”
50.04 - 50 = - = 2.67 0.015
and P(z > 2.67) = 0.5000 - 0.4962 = 0.0038. As the sample size increases, it is less likely that a mean value of at least 50.04 will be observed purely by chance. If it did, some special cause would likely be present. Confidence Intervals A confidence interval (Cl) is an interval estimate of a popula¬
tion parameter that also specifies the likelihood that the interval contains the true population parameter. This probability is called the level of confidence, denoted by 1 - a, and is usually expressed as a percentage. For example, we might state that "a 90 percent Cl for the mean is 10 ± 2." The value 10 is the point estimate calculated from the sample data, and 2 can be thought of as a margin for error. Thus, the interval estimate is [8,12]. However, this interval may or may not include the true population mean. If we take a different sample, we will most likely have a different point esti¬ mate, say 11.4, which determines the interval estimate [8.4,12.4]. Again, this interval may or may not include the true population mean. If we chose 100 samples, leading to 100 different interval estimates, we would expect that 90 percent of them—the level of confidence—would contain the true population mean. We would say we are 90 percent confident that the interval we obtain from sample data contains the true population mean. Commonly used confidence levels are 90, 95, and 99 percent; the higher the confidence level, the more assurance we have that the interval contains the true population parameter. As the confidence level increases, the confidence interval becomes larger to provide higher levels of assurance. Some common confidence intervals are • Confidence interval for the mean, standard deviation known, sample size = n: x ± zaPo/Vn • Confidence interval for the mean, standard deviation unknown, sample size = n: X ± fo/2,n-l(s/V«) • Confidence interval for a proportion, sample size = n: p ± z« y
I PQ-P) n
Chapter 11
Statistical Thinking and Applications
• Confidence interval for difference between two means, independent samples, equal variance, sample sizes = n1 and n2: X,-X2± fa/2,„, + „^2) Sp
r
i i~ — + —
V n,
n2
• Confidence interval for differences between two proportions, sample sizes = nt and n2:
p,-p2±za/2
+
P2(l~Pz) n2
Hypothesis Testing Hypothesis testing involves drawing inferences about two con¬ trasting propositions (hypotheses) relating to the value of a population parameter, one of which is assumed to be true in the absence of contradictory data. For instance, suppose that a company is testing out a prototype process that is designed to reduce manufacturing cycle time. They can evaluate the proposed process by testing a hypothesis that the mean cycle time is the same as the current process. A hypothesis test involves the following steps:
1. Formulate the hypotheses to test. 2. Select a level of significance that defines the risk of drawing an incorrect con¬ clusion about the assumed hypothesis that is actually true. 3. Determine a decision rule on which to base a conclusion. 4. Collect data and calculate a test statistic. 5. Apply the decision rule to the test statistic and draw a conclusion. To illustrate hypothesis testing, let us examine a producer of computer-aided design software for the aerospace industry that receives numerous calls for technical support. Tracking software is used to monitor response and resolution times. The company has a service standard of four days for the mean resolution time. However, the manager of the technical support group has been receiving some complaints of long resolution times. During one week, a sample of 44 customer calls resulted in a sample mean of 5.23 and standard deviation of 13.5. Even though the sample mean exceeds the four-day standard, does the manager have sufficient evidence to con¬ clude that the mean service time exceeds four days, or is this particular sample mean simply a result of sampling error? The hypothesis tested is H„: Mean response time < 4 H,: Mean response time > 4 The appropriate test statistic is x-4
t
S/Vn The decision rule is to reject Hn if t > t„_ha. We compute the value of the test statistic as X —4 1
~
s/VF
5.23-4 =
13.5/V44
1.24 =
2.035
= °-609'
Because f43i 05 = 1.6811, we cannot reject the null hypothesis. Therefore, the manager finds no sufficient statistical evidence that the mean response time exceeds 4.
541
542
Part 3
Six Sigma and the Technical System
Enumerative and Analytic Studies
One of the biggest mistakes that A static population, such as employees in a com¬ people make in using statistical pany or its customer base, can be aqalyzed to methods is confusing data that are sampled from a static population estimate population parameters such as the (cross-sectional data) with data mean, variance, or proportion. Confidence inter¬ sampled from a dynamic process vals and hypothesis tests can be applied. How¬ (time series data). ever, the purpose of sampling from a process is generally to predict the future. The characteris¬ tics of the population may change over time as a plot of sample means or variances might show. In such cases, confidence intervals and hypothesis tests are not appro¬ priate unless the time series can be shown to be stationary, that is, have a constant mean and variance over time. Examining a trend chart of the data over time can usually provide more stationary parameters. Deming called the analysis of a static popula¬ tion an enumerative study, and the analysis of a dynamic time series an analytic study. Applying classical statistical inferences to an analytic study is not appropriate because they provide no basis for prediction. Thus, it is important to understand how to apply statistical tools properly. In the hypothesis testing example, for instance, we need to assume that the data are stationary during the week over which they were collected to apply this tool cor¬ rectly. If we sampled the data over a long period of time and the characteristics of the population (mean or variance) changed over that time, then conducting a hypothesis test would not be appropriate. Design of Experiments Design of experiments (DOE), developed by R. A. Fisher in England, dates back to
the 1920s. A designed experiment is a test or series of tests that enables the experi¬ menter to compare two or more methods to determine which is better, or determine levels of controllable factors to optimize the yield of a process or minimize the vari¬ ability of a response variable.12 For example, a paint company might be interested in determining whether different additives have an effect on the drying time of paint in order to select the additive that results in the shortest drying time. As another example, suppose that two machines produce the same part. The material used in pro¬ cessing can be loaded onto the machines either manually or with an automatic device. The experimenter might wish to determine whether the type of machine and the type of loading process affect the number of defectives and then to select the machine type and loading process combination that minimizes the number of defectives. As a practical tool for quality improvement, experimental design methods have achieved considerable success in many industries. In a celebrated case, Ina Tile Com¬ pany, a Japanese ceramic tile manufacturer, had purchased a $2 million kiln from West Germany in 1953.13 Tiles were stacked inside the kiln and baked. Tiles toward the outside of the stack tended to have a different average size and more variation in dimensions than those further inside the stack. The obvious cause was the uneven temperatures inside the kiln. Temperature was an uncontrollable factor, a noise factor. To try to eliminate the effects of temperature would require redesign of the kiln itself, a very costly alternative. A group of engineers, chemists, and others who were familiar with the manufacturing process brainstormed and identified seven major controllable variables that could affect the tile dimensions: 1. Limestone content 2. Fineness of additive
Chapter 11
3. 4. 5. 6. 7.
Statistical Thinking and Applications
Content of agalmatolite Type of agalmatolite Raw material quantity Content of waste return Content of feldspar
The group designed and conducted an experiment using these factors. The experi¬ ment showed that the first factor, the limestone content, was the most significant factor, the other factors had smaller effects. By increasing the limestone content from 1 percent to 5 percent and choosing better levels for other factors, the percentage of size defects was reduced from 30 percent to less than 1 percent. Limestone was the cheapest material in the tile. In addition, the experiment revealed that a smaller amount of agalmatolite, the most expensive material in the tile, could be used without adversely affecting the tile dimension. Both the effect of the noise factor and the cost of the product were reduced at the same time! This discovery was a break¬ through in the ceramic tile industry. As another example, ITT Avionics Division, a leading producer of electronic war¬ fare systems, experienced a high defect rate when using a wave solder machine to solder assemblies on printed circuit boards.1-1 The wave solder machine, developed to eliminate hand soldering, transports printed circuit boards through a wave of solder under computer control. A brainstorming session identified 14 process variables. From three sets of designed experiments, the subsequent data resulted in decisions that lowered the defect rate from seven or eight to 1.5 per board. With 2,500 solder connections per board, this translated to a a defect rate of 600 defects per million con¬ nections. Another division, ITT's Suprenant Company, an electrical wire and cable manufacturer and a supplier to Ford, saved an estimated $100,000 per year in scrap, reduced product variability by a factor of 10, and improved the run rate of an extruding operation by 30 percent. Historically, experimental design was not widely used in industrial quality improvement studies because engineers had trouble working with the large number of variables and their interactions on many different levels in industrial problems. However, improved computer software and more sophisticated training have recently experimental design an important tool for quality improvement. Factorial Experiments One of the most common types of experimental designs is called a factorial experiment. In a factorial experiment, all combinations of levels of each factor are considered. For example, suppose that temperature and reaction time are identified as important factors in the yield of a chemical process. If the experiment is designed to analyze the effect of two levels of each factor (for instance, temperature at 100 and 120 degrees, and time at 60 and 75 minutes), then there would be 22 = 4 possible combinations to test: Temperature 100 degrees 100 degrees 120 degrees 120 degrees
Time 60 minutes 75 minutes 60 minutes 75 minutes
In general, an experiment with m factors at k levels would have k'" combinations. Each combination should be performed in a random fashion to eliminate any potential systematic bias. The purpose of a factorial experiment is to estimate the effects of each factor. For instance, what is the effect of a 20-degree change in temperature? Of a 15-minute
Part 3
544
Six Sigma and the Technical System
change in reaction time? These questions are answered easily by finding the differ¬ ences of the averages at each level. For instance, suppose we obtained the following results: Temperature 100 degrees 100 degrees 120 degrees 120 degrees
x Time 60 minutes 75 minutes 60 minutes 75 minutes
Yield (%) 85 88 90 80
The average yield for a temperature of 100 degrees is (85 + 88)/2 = 86.5. The average yield for a temperature of 120 degrees is (80 + 90)/2 = 85. Thus, the average difference in yield from increasing the temperature from 100 to 120 degrees is 85 - 86.5 = -1.5 per¬ cent. Similarly, the average difference in increasing the time from 60 to 75 minutes is (88 + 80)/2 - (85 + 90)/2 = 84 - 87.5 = -3.5 percent. These differences—how the changes in the level of one factor affect the response—are called main effects. Thus, we might conclude that increasing temperature or time decreases the yield of the process. However, in many situations, the effect of changing one factor depends on the level of other factors. For example, the effect of temperature may depend on the reac¬ tion time. In this example, we see that if the temperature is held constant at 100 degrees, an increase in reaction time results in a higher yield. However, when the temperature is 120 degrees, an increase in reaction time decreases the yield. This interaction is easy to determine by graphing the results as shown in Figure 11.15. If the lines are nearly parallel, then no interaction exists. In this case, we see an interaction. We may quantify the interaction by taking the average of difference of the yield when the temperature is increased from 100 to 120 degrees at a constant reaction time of 60, and subtracting the average of difference of the yield when the temperature is increased from 100 to 120 degrees at a constant reaction time of 75: Temperature x Time Interaction = (90 - 85)/2 - (80 - 88)/2 = 6.5 percent
Figure 11.15 Interaction Effects
Temperature
100
120
Chapter 11
Statistical Thinking and Applications
545
The closer this quantity is to zero, the smaller the interaction effect. In this case, a sig¬ nificant interaction is apparent. When interactions are present, main effects have little meaning; individual factors must be interpreted relative to levels of the other factors. We see that higher temperature and lower time appear to optimize the yield. The fol¬ lowing example shows a simple application of a three-factor experiment not unlike those used in industrial and business settings. Many one-tenth scale remote control (RC) model car racing enthusiasts believe that spending more money on high-quality batteries, using expensive gold-plated connectors, and storing batteries at low temperatures will improve battery life per¬ formance in a race. To test this hypothesis, an electrical test circuit was constructed to measure battery discharge under different configurations. Each factor (battery type, connector type, and temperature) was evaluated at two levels, resulting in 23 = 8 experimental conditions shown in Table 11.2. Calculations of the main effects are as follows: Battery cost Low = (72 + 93 + 75 + 94)/4 = 83.5 minutes High = (612 + 490 + 493 + 489)/4 = 521 minutes Main effect = High - Low = 437.5 minutes Connector type Gold-plated = (94 + 75 + 490 + 493)/4 = 288 minutes Standard = (72 + 93 + 612 + 489)/4 = 316.5 minutes Main effect = Standard - Gold-plated = 28.5 minutes Temperature Cold = (72 + 75 + 490 + 612)/4 = 312.25 minutes Ambient = (93 + 489 + 493 + 94)/4 = 292.25 minutes Main effect = Ambient - Cold = 20 minutes These results suggest that high cost batteries do have a longer life, but that the impacts of gold plating or battery temperature do not appear to be significant. Because only one factor appears to be significant, calculation of interaction effects are not required. These conclusions can be tested more rigorously using analysis of vari¬ ance, which will be discussed briefly next. Indeed, an analysis of variance confirms
Table 11.2 Experimental Design for Testing Battery Performance Experimental Run
Battery Type
Connector Type
Battery Temperature
Discharge Time (minutes)
1
High cost
Gold-plated
Ambient
493
2
High cost
Gold-plated
Cold
490
3
High cost
Standard
Ambient
489
4
High cost
Standard
Cold
612
5
Low cost
Gold-plated
Ambient
94
6
Low cost
Gold-plated
Cold
75
7
Low cost
Standard
Ambient
93
8
Low cost
Standard
Cold
72
546
Part 3
Six Sigma and the Technical System
that the battery cost factor is statistically significant while the other factors are indis¬ tinguishable from experimental error. Classical design of experiments cap require many, often costly, experimental runs to estimate all main effects and interactions. A Japanese engineer. Dr. Genichi Taguchi, proposed another approach to DOE. He developed an approach to designing experi¬ ments that focused on the critical factors while deemphasizing their interactions, which greatly reduced the number of required experiments. However, Taguchi's approach violates some traditional statistical principles and has been criticized bv the statistical community.16 To add to the shortcomings of his approach, Taguchi intro¬ duced some statistically invalid and misleading analyses, ignored modem graphical approaches to data analysis, and failed to advocate randomization in performing the experiments. Even though many of these issues are subject to debate, numerous com¬ panies have used Taguchi's approaches effectively. Analysis of Variance (ANOVA)
Because of Six Sigma, practitioners have "rediscovered" such techniques as analysis of variance, or ANOVA, which has long been an important tool in statistical analysis. ANOVA is a methodology for drawing conclusions about equality of means of mul¬ tiple populations. In its simplest form—one-way ANOVA—we are interested in com¬ paring means of observed responses of several different levels of a single factor. ANOVA tests the hypothesis that the means of all populations are equal against the alternative hypothesis that at least one mean differs from the others. To conduct an ANOVA, we need to 1. 2. 3. 4. 5.
Carefully define the purpose and assumptions of the experiment. Gather data related to the factor levels of interest. Compute ANOVA statistics. Interpret the meaning of the data. Take action.
Let us suppose that the model race car enthusiasts in the previous example were interested in determining whether any significant differences exist between various brands of batteries (the factor levels). Understanding possible differences in battery performance could be a first step in examining whether connection or temperature has an effect on performance. Table 11.3 shows discharge times for three different brands of batteries, gathered through a measurement process. Microsoft Excel provides a simple procedure to conduct a one-wav ANOVA. Select ANOVA: Single Factor from, the Tools/Data Analysis options. In the dialog box
Table 11.3 Battery Discharge Time Data by Brand Brand —
Observation
A
B
C
1
493
108
94
2
490
95
75
3
489
115
93
4
612
82
72
Chapter 11
Statistical Thinking and Applications
547
that pops up, enter the input range of the data in your spreadsheet and check whether it is stored in rows or columns. Table 11.4 shows the results of applying this tool. What does it tell us? The objective of ANOVA is to statistically test the differences between the means of the groups (the time to discharge for the various brands of batteries) to determine whether they are the same or at least one mean is different. To make this determina¬ tion, ANOVA partitions the total variability of the data into two parts, the variation between groups and the variation within groups. If the total variation between groups is relatively small compared to the variation within groups, it suggests that the populations are essentially the same. A relatively large variation between groups, however, suggests that differences exist in the unknown population means. The vari¬ ation in the data is computed as a sum of squared (SS) deviations from the appro¬ priate sample mean, and scaled as a variance measure, or "mean square" (MS). By dividing the mean square between groups by the mean square within groups, an F statistic is computed. If this value is larger than a critical value, Fcrit, then the data sug¬ gest that a difference in means exist. An examination of the SUMMARY section in Table 11.4 shows that Group A's mean value and variance are considerably larger than the others. In the ANOVA part of the table, the mean square between groups is significantly larger than the mean square within groups, resulting in an F statistic of 183.0412. When this value is com¬ pared to the critical F value (4.256), for 2 and 9 degrees of freedom at a 0.05 level of significance (from an F-table, available in any statistics text), we can reject the hypothesis that the means for the three battery types are the same. In fact, F = 183 is so much larger than 4.256, that we only have a 5.13 x 10-8 probability (the p-value in the output) that we could be wrong and should have failed to reject the hypothesis! The model race car enthusiasts could conclude that a significant difference exists between the battery types. Other statistical tests are available to demonstrate what
Table 11.4 Results of Microsoft Excel ANOVA Procedure Summary Groups
Count
Sum
Average
Variance
A
4
2084
521
3683.333
B
4
400
100
212.6667
C
4
334
83.5
135
ANOVA Source of Variation
Between Groups Within Groups
Total
SS
df
491892.67
2
12093
503985.67
9
11
MS
245946.33 1343.6667
F
183.0412
P-value
5.13E-08
F crit
4.256492
Part 3
548
Six Sigma and the Technical System
factor levels differ from the others (although in this case, it is fairly obvious). The next step might be to explore other variables (connector type, battery temperature) to see how they might affect battery discharge time. It would require more sophisticated ANOVA methods. We encourage you to consult more complete statistics books, such as Montgomery or Lipson and Sheth.17 You can probably identify many applications of ANOVA in Six Sigma projects, when differences among critical quality characteristics must be explored. However, ANOVA requires that some statistical assumptions be satisfied for proper interpreta¬ tion of the results, namely that the populations from which the samples are drawn are normally distributed and have equal variances, and that the data are randomly and independently obtained. These assumptions should be validated if possible. Regression and Correlation Regression analysis is a tool for building statistical models that characterize rela¬ tionships between a dependent variable and one or more independent variables, all of which are numerical. 'The relationship may be linear, one of many types of non¬ linear forms, or there may be no relationship at all. A regression model that involves a single independent variable is called simple regression. A regression model that involves several independent variables is called multiple regression. To develop a regression model, you first must specify the type of function that best describes the data. This step is important, because using a linear model for data that are clearly nonlinear, for instance, would probably lead to poor business decisions and results. The type of relationship can usually be seen in a scatter diagram, and we always rec¬ ommend that you create one first to gain some understanding of the nature of any potential relationship. Correlation is a measure of a linear relationship between two variables, X and Y, and is measured by the (population) correlation coefficient. Correlation coefficients will range from -1 to +1. A correlation of 0 indicates that the two variables have no linear relationship to each other. Thus, if one changes, we cannot reasonably predict what the other variable might do using a linear equation (we might, however, have a well-defined nonlinear relationship). A correlation coefficient of +1 indicates a perfect positive linear relationship; as one variable increases, the other will also increase. A correlation coefficient of-1 also shows a perfect linear relationship, except that as one variable increases, the other decreases. To illustrate regression, we will use a common issue in quality that we will dis¬ cuss again in a later chapter—ensuring that instruments are properly calibrated. In principle, it is a simple matter to check calibration. One connects the instrument to a known source, such as an extremely accurate voltage generator to check a voltmeter or a precision gauge block to check a micrometer. A reading is then obtained to deter¬ mine whether the instrument is capable of accurately measuring the known variable. In practice, numerous sources of variation in the process may make calibration difficult. The data in Figure 11.16 represent actual readings obtained from the calibration of a voltmeter versus the standard source readings from an accurate voltage generator. The source readings were purposely not set in even integer increments so as to mini¬ mize possible bias of the inspector taking the actual readings. To determine whether the instrument is accurate, we can develop a regression equation for the data. Using the Regression tool in Microsoft Excel, we obtain the results shown in Figure 11.16. The estimated regression equation is Y = 0.0265 + .9914X
Chapter 11
Statistical Thinking and Applications
549
1 2 3 4 5 6 / a 9 1U 11 12 Id 14 15 1b 1/ 18 iy
A Voltmeter Calibration
jo
Figure 11.16 Microsoft Excel Regression Results
D
F
E
G
H
1
J
SUMMARY OUTPUT Actual (Y) Source (X) 1.09 1.05 2.12 2.15 3.08 3.12 4.09 4.08 5.11 5.11 6.08 6.07 7.2 7.23 8 34 8.3 9.59 9.66 10.41 10,49
.. Regression Statistics Multiple R 0.999967043 R Square 0.999934087 Adjusted R Square 0.999925848 Standard Error 0.027282724 Observations 10
— -j . —
’. —
ANOVA df Regression Residual Total
Intercept Source (X)
1 8 9
SS MS 90.33725522 90.33726 0.005954776 0.000744 90.34321
Coefficients Standard Error 0.02648404 0.018447595 0.991364042 0.002845689
F Significance F 121364.4 5.16118E-18
t St at P-vaiue 1.435636 0.189028 348.374 5.16E-18
~
Lower 95% Upper 95% -0.016056219 0.069024299 0.984801867 0.997926217
The value of R2 is .9999, indicating an excellent fit. Note also that the value of the intercept is close to 0 and the slope is close to 1, which is where they should be. We would conclude that the instrument is in near perfect calibration, as the scatter chart in Figure 11.17 also indicates.
Figure 11.17 Scatter Chart of Voltmeter Calibration Readings
—
..| |
Part 3
550
Six Sigma and the Technical System
Quality in Practice Improving Quality of a Wave Soldering Process Through Design of Experiments18 A Printed Circuit Assembly-Encoder (PCAEncoder) is a critical component for the base car¬ riage assembly for a printer. The PCA-Encoder is produced by putting the electronic components on printed circuit boards (panels) that contain eight small boards, and then soldering the components using a wave soldering process. Any defect in any of the solder joints will lead to the failure of the circuit. Thus, it is important to ensure that soldering is defect-free. Typical soldering defects are blowholes (insufficient solder) and bridges (solder between two joints). At a Hewlett-Packard India Ltd. plant in Bangalore, India, a high level of soldering defects was observed, necessitating 100 percent inspection for all circuit boards. Any defects identified required manual rework, which consumed much time. A study was undertaken to optimize the wave soldering process for reducing defects, thereby eliminating the inspection stage after the process. The quality engineers conducted a detailed study on the solder defects to understand what aspects of the wave soldering process might affect the resulting quality. These were identified as 1. 2. 3. 4. 5. 6. 7. 8.
Conveyor speed Conveyor angle Solder bath temperature Solder wave height Vibration of wave Preheater temperature Air knife Acid number (solid content in the flux), which is difficult to control because of envi¬ ronmental conditions
The engineers decided to use experimental design because of the long time frame required to adjust process parameters by trial and error, and the lack of insight into the possible joint effects of different parameters. Based on discussions with technical personnel, seven factors at three levels were selected for the experiment, as shown in Table 11.5. Conveyor speed and conveyor angle were fixed. A full factorial experiment would take 1,458 trials to conduct, which was not deemed to be practical. From statistical theory in the design of experi¬ ments, the seven main effects could be estimated by conducting only 18 trials as shown in Table 11.6.
Table 11.5 Factors and Levels for Experimenta¬
tion Level Factor Bath tem¬ perature fC) Wave height" Overheated preheater (OH-PH) (PH-1) Preheater-1 (PH-1) CC) Preheater (PH-2) CC) Air Knife Omega'
Code
1
2
3
A B
248* 4.38
252 4.40*
4.42
C
340
360*
380
D
340
360*
380
E
340
F
G
360* 3* 2*
380 6 4
“Existing level "The wave height is measured as the rpm of the motor pumping the solder. “Omega refers to the vibration of the solder wave.
The experimental outcomes (response) were the number of defective solder joints in a frame (352 joints). Each experiment was repeated three times. Using analysis of variance, it was observed that bath temperature, wave height, and omega had a significant effect on the soldering defects. By setting the factors at the optimum levels identified through the experiments, the predicted defect level was 1670 ppm as opposed to the current rate of more than 6000 ppm. However, the predicted average and the result of a confirmatory experi¬ ment were not sufficient to eliminate inspection completely, so additional experimental designs were conducted to reduce defects. The next experiment considered the results of the first experiment and some of the uncontrol¬ lable factors. However, the different levels of the significant factors from the first experiment were selected in such a way that the new levels were allowed to vary around the optimum level of the first experiment. Based on the results of these additional experiments, new optimum levels of factors were identified and implemented with sig¬ nificant improvements. Figure 11.18 shows the
Chapter 11
Statistical Thinking and Applications
551
Table 11.6 Data Corresponding to the First Experiment
Exp. No.
(1) Bath Temp. ( C)
(2) Wave Height
(3) OH-PH
(4) PH-2
ra
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
248 248 248 248 248 248 248 248 248 252 252 252 252 252 252 252 252 252
4.38 4.38 4.38 4.40 4.40 4.40 4.42 4.42 4.42 4.38 4.38 4.38 4.40 4.40 4.40 4.42 4.42 4.42
340 360 380 340 360 380 340 360 380 340 360 380 340 360 380 340 360 380
rci
(5) PH-1 (°C)
(6) Air Knife
(7) Omega
340 360 380 340 360 380 360 380 340 380 340 360 360 380 340 380 340 360
340 360 380 360 380 340 340 360 380 380 340 360 380 340 360 360 380 340
0 3 6 3 6 0 6 0 3 3 6 0 0 3 6 6 0 3
0 2 4 4 0 2 2 4 0 2 4 0 4 0 2 0 2 4
Figure 11.18 Solder Defects After Experimental Design Optimization
Days
1 1 0 0 1 4 8 2 4 2 1
1 6 3 4 2 2 2 4
Response 2 3 2 2 1 0 2 1 4 1 2 3 2 3 3 3 1
7 1 2
1 0 0 1 0 6 3 0 4 1 1 2 4 8 1 3 3 1
Part 3
552
ppm level during the course of the experimenta¬ tion, which took only 45 days. Key Issues for Discussion
1. Why did the first experimental design not find the true optimum combination of factors
Six Sigma and the Technical System
to achieve the maximum reduction of defects? 2. What were some of the advantages of using experimental design over a traditional trialand-error approach?
Quality in Practice Applying Statistical Analysis in a Six Sigma Project at (This Quality in Practice features another project performed at the GE Fanuc plant. We encourage you to read the related case in Chapter 10 for some additional background.) In mid-2002 a team at the GE Fanuc manufac¬ turing plant in Charlottesville, Virginia, led by Six Sigma Black Belt Donald Splaun was given the goahead to investigate Black Belt Project #P52320. The objective of the project was to evaluate Printed Wire Board (PWB) Fabricated Board Finishes to determine if the high-priced nickel-gold (Ni-Au) finished boards that were being used were neces¬ sary as mounting platforms for fine pitch surfacemounted devices (SMDs) or for fine-pitched Ball Grid Array (BGA) electronic controller boards. SMDs are electronic components, such as micro¬ processors, that are placed on the top of electronic circuit boards (fabricated boards) and then have their electrical wire leads soldered into place. Fine pitch SMDs don't have much space between their electrical wire leads, making it difficult to put just the right amount of solder on them to make the proper electrical connect to the circuit boards on which they are mounted. The completed boards with all components properly mounted on them are then used in electrical assemblies to control the operations of industrial machinery. Splaun had seven people on his analysis team, plus a financial representative to verify the dollar costs and savings; a Master Black Belt reviewer, who would evaluate the project to prevent obvious gaps in the analysis; and perhaps most importantly from a managerial standpoint, a Champion/Sponsor, who would ensure project visibility and that resources were allocated to com¬ plete the project. Team members and their job functions were:
GE
Fanuc19
Team Members Titles and Functions
Team Leader and Black Belt Process engineer and fabricated board expert Advanced manufacturing engineer responsible for SMD board assembly Sourcing agent who purchases the boards Test engineer who tests and evaluates boards Producibility engineer who works with design teams Production line operator who runs boards Supplier quality analysis technician responsible for incoming board quality Outside Resources
Financial representative to assist in cost calcula¬ tions Champion / Sponsor Reviewer and Master Black Belt After being formed, the team used a 12-step DMAIC process developed by GE Fanuc, to guide them through the project (see Table 11.7). The first three pre-project definition substeps (A, B, and C) required them to identify project CTQs, develop a team charter and have it approved, and define a process map. Project team members identified the CTQs by using a standard cause-and-effect matrix, and weighted rankings of CTQ factors to prioritize them. The team determined there were three CTQs, two of which were business factors and the other a project factor: • Business CTQ factors: Variable cost produc¬ tivity (VCP) improvement, composed of the CTQs of internal cost reduction and contri¬ bution margin improvement • Project CTQ factor: Benefits associated with
Chapter 11
Statistical Thinking and Applications
553
Table 11.7 GE Fanuc 12-Step DMAIC Process Step
Description
Tools
Deliverables
Define A
Identify project CTQs
Project CTQs (1)
B
Develop team charter
Approved charter (2)
C
Define process map
High-level process map (3)
Measure 1
Select CTQ characteristics
Customer, QFD, FMEA
Project Y (4)
2
Define performance standards
Customer, Blueprints
Performance standard for Project Y (5)
Measurement system analysis
Continuous Gage R&R,
Data collection plan and
Test/Retest, Attribute R&R
MSA (6), Data for Project Y (7) Process capability—Project Y (8)
Analyze
4
Establish process capability
Capability indices
5
Define performance objectives
Team, Benchmarking
Improvement goal for Project Y (9)
6
Identify variation sources
process and graphical
Prioritized list of all Xs (10)
Analysis, Flypothesis tests Improve
7 8 9
Screen potential causes
DOE-screening
List of vital few Xs (11)
Discover variable relationships
Factorial designs
Proposed solution (12)
Establish operating tolerances
Simulation
Piloted solution (13)
Control 10
Define and validate measurement
Continuous Gage R&R, MSA
system on Xs (independent
Test/Retest, Attribute R&R
variables) in actual application 11
Determine process capability
Capability indices
Process capability Y, X (14)
12
Implement process control
Control charts,
Sustained solution (15),
Mistake proof, FMEA
Documentation (16)
additional Ni-Au cost of $190,000 per year, suspected to be unnecessary The team developed their charter to define the problem and working relationships. The problem clearly and succinctly stated as: GE Fanuc currently specifies Ni-Au on fine-pitch SMD and BGA boards. The purpose of this project is to evaluate if this specifica¬ tion is necessary. The team also identified tools and databases that were to be used in the study, not only to ensure that everyone was working from a
common source, but also to take advantage of the training that had been provided to team members. The tools included two statistical/spreadsheet software packages (Minitab and Excel) and a plantwide integrated database (SAP) that con¬ tained information on board characteristics, usage, specifications, costs, and so on. Based on a 29-step process flowchart, it was decided that the analysis would require the use of a moderately complex experimental design. This design was required to determine the effects of supplier differences and finishes, because relatively
554
few defects were being observed in manufacturing the boards. Data would have to be gathered from the experiment and from supplier surveys to help the team track potential causes that could have a bearing on the functionality and cost of each of the alternative boards or board materials being considered. The experiment was designed to sample and test 288 CX3A1 boards: • 96 hot air solder leveled (HASL) boards, 32 from each supplier • 96 nickel-gold (Ni-Au) boards, 32 from each supplier • 96 silver (Ag) boards, 32 from each supplier and to evaluate three suppliers. The three suppliers were: • Vendor G, Singapore/China (one or two GE Fanuc production suppliers) • Vendor P, Taiwan/China (second major GE Fanuc supplier) • Vendor D, USA (current prototype/fast turn supplier) The cause-and-effect matrix identified 13 characteristics (Xs, or independent variables) that were considered important to measure during the experiment for each of the three finish types (Ys, or dependent variables). The primary hypothesis was that no significant differences in numbers of defects would be incurred, regardless of finish. In addition, a hypothesis that no significant interaction effects existed between suppliers, coatings, and any of the 13 characteristics considered essential for quality board functioning was investigated. The data col¬ lection and analysis process consisted of eight care¬ fully defined steps conducted over a six-day period, involving almost $ 37,500 worth of boards and hard-to-measure production delays while the test boards were run on what is normally high speed, highly automated production machines. After the data were collected, numerous ANOVA computer runs were made to pinpoint problem areas and test hypotheses. It was espe¬ cially important to test the capabilities of each of the three types of board finishes to determine whether they were equivalent to the current, and very expensive, Ni-AU finished boards. It was also necessary to get some data to prove or disprove
Part 3
Six Sigma and the Technical System
hypotheses about supplier capabilities as well. Table 11.8 shows a typical computer printout and analysis of one of the 13 variables that was tested, called "Wave Solder Skips." Of 15 ANOVA analysis runs performed on the 13 experimental variables that were measured, eight showed no significance, primarily because those variables had zero defects. Other findings included: • Ni-Au boards are not significantly different from or better than horizontally processed HASL or Silver Boards for fine pitch SMD processing, therefore the firm can save money by switching from Ni-Au to HASL or Silver. • The company should not use Vendor D for production. Results and suggestions for improvement of their prototype quality should be discussed with them. • Ni-Au is worse for wave soldering, based on a defect measure of "insufficient solder fill." • Vendor G was found to have an issue with a defect measure of "GR False Failures" (to be reviewed with the supplier). • The GE Fanuc PWB Fab Specifications should be changed to reflect these conclusions. From these analyses, the summary conclusion was that GE Fanuc did not need nickel-gold boards for fine pitch SMD. The estimated savings from this project ranged from 7.1 percent on two-layer boards to 22.8 percent on four-layer boards, with an average of 14.3 percent savings on these 89 board types, and total estimated savings of $190,000 per year. Key Issues for Discussion
1. Why did the experimental design have to be so complex? Why were so many individuals involved in this project? 2. What might have been some contributing factors that caused the company to select the Ni-Au over the cheaper boards in the past? 3. For Table 11.8, what can you conclude, given the F values and the p-values in the table? What steps should the team take, regarding use of vendors and further testing for this particular independent variable?
Chapter 11
Statistical Thinking and Applications
555
Table 11.8 Typical ANOVA Output for Vendor and Finish Analysis Solder Skips, Analysis All Two-Way Analysis of Variance Analysis of Variance for Wave Solder Skips Source
DF
SS
MS
F
P
Vendors
2
36.55
18.27
15.27
0.000
Finish
2
16.44
8.22
6.87
0.001
Interaction
4
23.39
5.85
4.89
0.001
1.20
Error
279
333.94
Total
287
410.32
Process Variable Averages Manufacturers Mean Number of Defects Vendor C Vendor D Vendor G
0.03 0.80 0.06
Finish HASL Ni-Au Silver
0.271 0.021 0.604
Review Questions 1. What is statistical thinking? Why is it important to managers and workers at all levels of an organization? 2. Explain the difference between common and special causes of variation. 3. Explain the two fundamental mistakes that managers make when attempting to improve a process. Can you cite any examples in your personal experience in which such mistakes were made? 4. What are the lessons of the Red Bead and Funnel Experiments? Can you cite any examples in your experience where someone acted counter to these lessons? 5. Discuss the differences among the three major components of statistical methodology (descriptive statistics, statistical inference, and predictive statis¬ tics). Why might this distinction be important to a manager? 6. Provide some examples of discrete and continuous random variables in a quality management context. 7. Define a population and a sample. What are their major characteristics? 8. Explain the difference between the standard deviation and the standard error of the mean. How are they related?
Part 3
Six Sigma and the Technical System
9. State the meaning of the central limit theorem in your own words. How impor¬ tant is it to the development and use of statistical quality control techniques? 10. What two factors influence sampling procedures? 11. Discuss the basic questions that must be addressed in a sampling study. 12. Describe the different methods of sample selection and provide an example in which each would be most appropriate. 13. What are the sources of systematic error in sampling? How can systematic error be overcome? 14. What is the purpose of design of experiments? 15. Describe a factorial experiment. Provide some examples of factorial experi¬ ments that you might use to solve some type of quality-related problem. 16. What limitations of simple factorial experiments can be overcome by using ANOVA? 17. What is the statistical basis for ANOVA; that is, what is it designed to test, sta¬ tistically?
Problems Note: Data sets for several problems in this chapter are available in the Excel work¬ book CllData.xls on the CD-ROM accompanying this text. Click on the appropriate worksheet tab as noted in the problem (e.g., Prob. 11-1) to access the data. 1. Use the data for Staunton Steam Laundry for the weights of loads of clothes processed through their washing department in a week. (See Prob. 11-1 in CllData.xls for data). Apply the Descriptive Statistics and Histogram tools in Excel to compute the mean, standard deviation, and other relevant statistics, as well as a frequency distribution and histogram for the following data. From what type of distribution might you suspect the data are drawn? 2. Apply the Descriptive Statistics and Histogram analysis tools in Excel to compute the mean, standard deviation, and other relevant statistics, as well as a frequency distribution and histogram for the following data (Prob. 11-2 in CllData.xls). From what type of distribution might you suspect the data are drawn? 3. The data (Prob. 11-3 in CllData.xls) represent the weight of castings (in kilo¬ grams) being made in the Harrison Metalwork foundry. Based on this sample of 100 castings, compute the mean, standard deviation, and other relevant statis¬ tics, as well as a frequency distribution and histogram. What do you conclude from your analysis? 4. The data (Prob. 11-4 in CllData.xls) show the weight of castings (in kilograms) being made in the Harrison Metalwork foundry after process changes took place. Compute the mean, standard deviation, and other relevant statistics, as well as a frequency distribution and histogram. Based on this sample of 100 castings, what do you conclude from your analysis? 5. Georgia Tea is sold in 2 liter (2000 milliliter) bottles. The mean volume of tea in the bottle is 2 liters and the standard deviation is 15 milliliters. If the process requires a 2 percent (total) or less probability of over- or underfilling, what should the upper and lower fill limits be? 6. New Orleans Punch is sold in 16-ounce cans. The mean number of ounces placed in a can is 15.8 with a standard deviation of 0.1 ounce. Assuming a normal distribution, what is the probability that the filling machine will cause
Chapter 11
Statistical Thinking and Applications
557
an overflow in a can, that is, the probability that more than 16 ounces will be placed in the can? 7. Kiwi Blend is sold in 950 milliliter (ml) cans. The mean volume of juice placed in a can is 945 ml with a standard deviation of 15 ml. Assuming a normal distri¬ bution, what is the probability that the filling machine will cause an overflow in a can, that is, the probability that more than 945 ml will be placed in the can? 8. Outback Beer bottles have been found to have a standard deviation of 5 ml. If 5 percent of the bottles contain less than 535 ml, what is the average filling volume of the bottles? 9. The standard deviation of the weight of filled salt containers is 0.4 ounce. If 2.5 percent of the containers contain less than 16 ounces, what is the mean filling weight of the containers? 10. In filling bottles of L & E Cola, the average amount of overfilling should be kept as low as possible. If the mean fill volume is 12.1 ounces and the standard devi¬ ation is 0.05 ounce, what percentage of bottles will have less than 12 ounces? More than 12.1 ounces (assuming no overflow)? 11. In a filling line at E & L Foods, Ltd., the mean fill volume for rice bubbles is 480 grams and the standard deviation is 3 grams. What percentage of containers will have less than 475 grams? More than 486 grams (assuming no overflow)? 12. The frequency table that follows represents the weight of castings (in kilo¬ grams) being made in the Harrison Metalwork foundry (see Problem 11-3 in CllData.xls for "raw" data). a. Based on this sample of 100 castings, find the mean and standard deviation of the sample. (Note: If only the given data are used, it will be necessary to research formulae for calculating the mean and standard deviations using grouped data from a statistics text.) b. Use an Excel spreadsheet if not already done for Problem 3, to plot the his¬ togram for the data. c. Plot the data on normal probability paper to determine whether the distrib¬ ution of the data is approximately normal. (Note: The Regression tool in Excel has a normal probability plot that may be used here.) Frequency Table
Cell Cell Cell Cell Cell Cell Cell Cell Cell
1 2 3 4 5 6 7 8 9
Upper Cell Boundaries
Frequencies
Cumulative %
37.5 37.8 38.1 38.4 38.7 39.0 39.3 39.6 39.9
1 3 8 26 29 15 13 4 1
1.00% 4.00% 12.00% 38.00% 67.00% 82.00% 95.00% 99.00% 100.00%
13. The frequency table that follows shows the weight of castings (in kilograms) being made in the Harrison Metalwork foundry after process changes took place (see Problem 11-4, in CllData.xls, for "raw" data). a. Based on this sample of 100 castings, find the mean and standard deviation of the sample. (Note: If only the given data are used, it will be necessary to research formulae for calculating the mean and standard deviations using grouped data from a statistics text.)
Part 3
558
Six Sigma and the Technical System
b. Use an Excel spreadsheet, if not already done for Problem 4, to plot the his¬ togram for the data. c. Plot the data on normal probability paper to determine whether the distrib¬ ution of the data is approximately normal. (Note: The Regression tool in Excel has a normal probability plot that may be used here.) Frequency Table
Cell Cell Cell Cell Cell Cell Cell Cell Cell
1 2 3 4 5 6 7 8 9
Upper Cell Boundaries
Frequencies
Cumulative %
37.5 37.8 38.1 38.4 38.7 39.0 39.3 39.6 39.9
1 3 8 23 25 23 10 6 1
1.00% 4.00% 12.00% 35.00% 60.00% 83.00% 93.00% 99.00% 100.00%
14. A utility requires service operators to answer telephone calls from customers in an average time of 0.1 minute or less. A sample of 30 actual operator times was drawn, and the results are given in the following table. In addition, operators are expected to determine customer needs and either respond to them or refer the customer to the proper department within 0.5 minute. Another sample of 30 times was taken for this job component and is also given in the table. If these variables can be considered to be independent, is the average time taken to per¬ form each component statistically different from the standard? Component
Mean Time
Standard Deviation
Answer Service
0.1023 0.5290
0.0183 0.0902
Note: Problems 15-19 address sample size determination and refer to theory covered in the Bonus Material for this chapter as contained on the student CD-ROM. 15. Determine the appropriate sample size to estimate the proportion of sorting errors at a post office at a 95 percent confidence level. Historically, the sorting error rate is 0.025, and you wish to have an allowable statistical error of 0.01. 16. You are asked by a motel owner to develop a customer satisfaction survey to determine the percentage of customers who are dissatisfied with service. In the past year, 20,000 customers were serviced. He desires a 95 percent level of con¬ fidence with an allowable statistical error of ±- 0.02. From past estimates, the manager believes that about 7 percent of customers have expressed dissatisfac¬ tion. What sample size should you use for this survey? 17. A local telephone company interviewed 150 customers to determine their satisfac¬ tion with service. Twenty-seven expressed dissatisfaction. Compute a 90 percent confidence interval for the proportion satisfied with an allowable error of 0.05. 18. A management engineer at Country Squire Hospital determined that she needs to take a work sampling study to see whether the proportion of idle time in the diagnostic imaging department had changed since being measured in a pre¬ vious study several years ago. At that time, the percentage of idle time was 10 percent. If the engineer can only take a sample of 800 observations due to cost factors, and can tolerate an allowable error of 0.02, what percent confidence level can be obtained from the study?
Chapter 11
Statistical Thinking and Applications
559
19. Using the Discovery Sampling table in the bonus materials, suppose that a pop¬ ulation consists of 2,000 units. The critical rate of occurrence is 1 percent, and you wish to be 99 percent confident of finding at least one nonconformity. What sample size should you select? 20. A process engineer at Sival Electronics is trying to determine whether a newer, more costly design involving a gold alloy in a computer chip is more effective than the present, less expensive silicon design. She wants to obtain an effective output voltage at both high and low temperatures, when tested with high and low signal strength. She hypothesizes that high signal strength will result in higher voltage output, low temperature will result in higher output, and the gold alloy will result in higher output than the silicon material. She hopes that the main and interaction effects with the expensive gold will be minimal. The following data were gathered in testing of all 2" combinations. What recom¬ mendation would you make, based on these data? Signal
Material
Temperature
Output Voltage
High High High High Low Low Low Low
Gold Gold Silicon Silicon Gold Gold Silicon Silicon
Low High Low High Low High Low High
18 12 16 10
8 11 7 14
21. The process engineer at Sival Electronics was trying to determine whether three suppliers would be equally capable of supplying the mounting boards for the new gold plated components that she was testing. The following table shows the coded defect levels for the suppliers, according to the finishes that were tested. Lower defect levels are preferable to higher levels. Using one-way ANOVA, ana¬ lyze these results. What conclusion can be reached, based on these data? Supplier 1 Finish Finish Finish Finish Finish
1 2 3 4 5
11.9 10.3 9.5 8.7 14.2
Supplier 2 6.8 5.9 8.1 7.2 7.6
Supplier 3 13.5 10.9 12.3 14.5 12.9
Projects, Etc. 1. A computer version of Deming's Funnel Exercise is available (free) at http:// www.qualitystation.com/Funnel-Free.htm. Download and run the funnel sim¬ ulation. Does it simulate the same rules as described in this chapter? 2. Devise an experiment similar to the battery performance test example to test different levels of some factor and conduct a statistical analysis of the results. Write up your experiment and results in a report along with the conclusions that you reach from the analysis. You might wish to consult the following paper: "101 Ways to Design an Experiment, or Some Ideas About Teaching Design of Experiments" by William G. Hunter, Technical Report No. 413 dated
Part 3
560
Six Sigma and the Technical System
June 1975 at http://www.stat.wisc.edu/department/handouts/technical413/ technical413.html for some ideas. 3. Using one sheet of paper, design and build a helicopter. Some methods of making a paper helicopter can be found at: http://www.exploratorium. edu/science_explorer/roto-copter.html and http://www.faa.gov/education/ resource/helicopt.htm. Use design of experiments to develop a design that keeps the helicopter air¬ borne for as long as possible.
KG»* I. The Disciplinary Citation20
A local delivery service has 40 drivers who deliver packages throughout the metropolitan area. Occa¬ sionally, drivers make mistakes, such as entering the wrong package number on a shipping docu¬ ment, failing to get a signature, and so on. A total of 240 mistakes were made in one year as shown in Table 11.9. The manager in charge of this opera¬ tion has issued disciplinary citations to drivers for each mistake.
Discussion Questions
1. What is your opinion of the manager's approach? How does it compare with the Deming philosophy? 2. How might the analysis of these data help the manager to understand the variation in the system? (Plot the data to obtain some insight.) How can the data help the manager to improve the performance of this system?
Table 11.9 Delivery Driver Citations Driver No. Mistakes
1 6
2 1
3 0
4 14
5 0
6 2
7 18
8 2
9 5
10 13
11 1
12 4
13 6
14
Driver No. Mistakes
15 0
16 0
17 1
18 3
19 15
20 24
21 3
22 4
23 1
24 2
25 3
26 22
27 4
28 8
Driver No. Mistakes
29 2
30 6
31 8
32 0
33 9
34 20
35 9
36 0
37 3
38 14
39 1
40 1
II.
5
The Quarterly Sales Report21
Suppose that Ron Hagler, the vice president of sales for Selit Corp., had just gotten a report on the past five years of quarterly sales data for the regions under his authority (see Table 11.10). Not happy with the results, he got on the phone to his secretary. "Marsha, tell the regional managers I need to speak with them this afternoon. Everyone must attend."
Marsha had been Hagler's secretary for almost a decade. She knew by the tone in his voice that he meant business, so she contacted the regional managers about the impromptu meeting at 2 p.m. At 1:55 p.m., the regional managers filed into the room. The only time they were called into a meeting together was when Hagler was unhappy. Hagler wasted no time. "I just received the
Chapter 11
Statistical Thinking and Applications
561
Table 11.10 Five Years of Sales Data by Region (Data in thousands) 1999 Sales Region Northeast
First Quarter $
Second Quarter
Third Quarter
Fourth Quarter
924
$ 928
$ 956
Southwest
1,056
1,048
1,129
1,073
Northwest
1,412
1,280
1,129
1,181
North Central
431
470
439
431
Mid-Atlantic
539
558
591
556
South Central
397
391
414
407
Second Quarter
Third Quarter
$1,222
2000 Sales Region Northeast
First Quarter
Fourth Quarter
$ 748
$ 962
$ 983
Southwest
1,157
1,146
1,064
1,213
Northwest
1,149
1,248
1,103
1,021
$1,024
North Central
471
496
506
573
Mid-Atlantic
540
590
606
643
South Central
415
442
384
448
2001 Sales Region
First Quarter
Second Quarter
Third Quarter
Fourth Quarter
$ 991
978
1,040
$1,295
Southwest
1,088
4,322
1,256
1,132
Northwest
1,085
1,125
910
999
403
440
371
405
Northeast
North Central Mid-Atlantic
657
602
596
640
South Central
441
366
470
426
Second Quarter
Third Quarter
2002 Sales Region
First Quarter
Fourth Quarter
Northeast
$ 756
$1,008
$1,038
$ 952
Southwest
4,352
1,353
1,466
1,196
Northwest
883
851
997
878
North Central
466
536
551
670
Mid-Atlantic
691
723
701
802
South Central
445
455
363
462
Second Quarter
Third Quarter
2003 Sales Region Northeast
First Quarter
Fourth Quarter
$1,041
$1,020
$ 976
$1,148
Southwest
1,330
1,003
1,197
1,337
Northwest
939
834
688
806
North Central
588
699
743
702
Mid-Atlantic
749
762
807
781
South Central
420
454
447
359
562
quarterly sales report. Northeast sales were fan¬ tastic. Steve, you not only improved 17.6 percent in the fourth quarter, but you also increased sales a whopping 20.6 percent over the previous year. I don't know how you do it!" Steve smiled. His phi¬ losophy to end the year with a bang by getting customers to stockpile units had paid off again. Hagler had failed to notice that Steve's first quarter sales were always sluggish. Hagler continued: "Terry, Southwest sales were also superb. You showed an 11.7 percent increase in the fourth quarter and an 11.8 percent increase over the previous year." Terry also smiled. She wasn't sure how she did so well, but she sure wasn't going to change anything. "Jan, Northwest sales were up 17.2 percent in the fourth quarter, but down 8.2 percent from the previous year," said Hagler. "You need to find out what you did previously to make your sales go through the roof. Even so, your performance in the fourth quarter was good." Jan tried to hide his puzzlement. Although he had received a big order in November, it was the first big order he had received in a long time. Overall, sales for the Northwest were declining. Hagler was now ready to deal with the "problem" regions. "Leslie, North Central sales were down 5.5 percent in the fourth quarter, but up 4.7 percent from the previous year. I don't understand how your sales vary so much. Do you
Part 3
Six Sigma and the Technical System
need more incentive?" Leslie looked down. She had been working very hard the past five years and had acquired numerous new accounts. In fact, she received a bonus for acquiring the most new business in 1998. "Kim, Mid-Atlantic sales were down 3.2 per¬ cent in the fourth quarter and down 2.6 percent from the previous year. I'm very disappointed in your performance. You were once my best sales representative. I had high expectations for you. Now, 1 can only hope that your first quarter results show some sign of life." Kim felt her face get red. She knew she had sold more units in 2003 than in 2002. "What does Hagler know anyway," she thought to herself. "He's just an empty' suit." Hagler turned to Dave, who felt a surge of adrenaline. "Dave, South Central sales were the worst of all! Sales were down 19.7 percent in the fourth quarter and down 22.3 percent from the previous year. How can you explain this? Do you value your job? I want to see a dramatic improve¬ ment in this quarter's results or else!" Dave felt numb. It was a tough region, with a lot of competi¬ tion. Sure, accounts were lost over the years, but those lost were always replaced with new ones. How could he be doing so badly? How can Ron improve his approach by applying principles of statistical thinking? Use any analyses of the data that you feel are appropriate to fully explain your thinking and help him.
III. The HMO Pharmacy Crisis22
John Dover had just completed an intensive course, "Statistical Thinking for Continuous Improvement," that was offered to all employees of a large health maintenance organization (HMO). There was no time to celebrate, however. Dover worked as a phar¬ macy assistant in the HMO's pharmacy and he was under a lot of pressure, because his manager, Juan de Pacotilla, was about to be fired. Pacotilla's dis¬ missal appeared imminent because of numerous complaints and even a few lawsuits over inaccurate prescriptions. Pacotilla now was asking Dover for his assistance in hying to resolve the problem. "John, I really need your help," said Pacotilla. "If I can't show some major improvement or at least a solid plan by next month, I'm history." "I'll be glad to help," replied Dover, "but what can I do? I'm just a pharmacy assistant."
"Your job title isn't important. I think you're just the person who can get this done," said Pacotilla. "I realize that I've been too far removed from day-to-day operations in the pharmacy, but you work there every day. You're in a much better position to find out how to fix the problem. Just tell me what to do, and I'll do it." "But what about the statistical consultant you hired to analyze the data on inaccurate prescrip¬ tions?" asked Dover. "To be honest. I'm really disappointed with that guy. He has spent two weeks trying to come up with a new modeling approach to predict weekly inaccu¬ rate prescriptions. I tried to explain to him that I don't want to predict the mistakes; I want to elimi¬ nate them. I don't think I got through, however, because he said we need a month of additional data
Chapter 11
Statistical Thinking and Applications
to verify the model before he can apply a new method he just read about in a journal to identify 'change points in the time series/ whatever that means. But get this, he will only identify the change points and send me a list. He says it's my job to figure out what they mean and how to respond. I don't know much about statistics. The only thing I remember from my course in college is that it was the worst course I ever took. I'm becoming con¬ vinced that statistics really doesn't have much to offer in solving real problems. Since you've just gone through the statistical thinking course, maybe you can see something I can't. I realize it's a long shot, but I was hoping you could use this as the pro¬ ject you need to officially complete the course." "I used to feel the same way about statistics, too," replied Dover. "But the statistical thinking course was interesting because it didn't focus on crunching numbers. I have some ideas about how we can approach making improvements in pre¬ scription accuracy. I think it would be a great pro¬ ject. But we might not be able to solve this
563 problem ourselves. As you know, there is a lot of finger pointing going on. Pharmacists blame the doctors' sloppy handwriting and incomplete instructions for the problem. Doctors blame the pharmacy assistants, who do most of the com¬ puter entry of the prescriptions, claiming that they are incompetent. Pharmacy assistants blame the pharmacists for assuming too much about their knowledge of medical terminology, brand names, known drug interactions, and so on." "It sounds like there's no hope," said Pacotilla. "I wouldn't say that at all," replied Dover. "It's just that there might be no quick fix we can do by ourselves in the pharmacy. Let me explain what I'm thinking about doing and how I would pro¬ pose attacking the problem using what I just learned in the statistical thinking course." How do you think John should approach this problem, using what he has just learned? Assume that he really did pick up a solid understanding of the concepts and tools of statistical thinking in the course.
ENDNOTES 1. Adapted from Brian L. Joiner, Fourth Generation Management (New York: McGraw-Hill, 1994), 129. 2. J. M. Juran and Frank M. Gryna, Jr., Quality Plan¬ ning and Analysis, 2d ed. (New York: McGraw-Hill, 1980), 35. 3. Scott M. Paton, "Juran: A Lifetime of Quality: An Exclusive Interview with a Quality Legend," Quality Digest, August 2002,19-23. 4. Adapted from Galen Britz, Don Emerling, Lynne Hare, Roger Hoerl, and Janice Shade, "How to Teach Others to Apply Statistical Thinking," Quality Progress, June 1997, 67-79. © 1997, American Society for Quality. Reprinted with permission. 5. Kimberly Weisul, "So Your Lie May Always Be True," Business Week, February 25, 2002,16. 6. Steven A. Melnyk and R. T. Christensen, "Vari¬ ance is Evil," APICS The Performance Advantage, June 2002,19. 7. Ronald D. Snee, "Getting Better Business Results: Using Statistical Thinking and Methods to Shape the Bottom Line," Quality Progress, June 1998,102-106. 8. Britz, et al. (see note 4). 9. Based on descriptions given in W. Edwards Deming, The Nezv Economics For Industry, Government, Education (Cambridge, MA: MIT Center for Advanced Engineering Study, 1993). 10. Frank H. Squires, "The Triumph of Statistics," Quality, February 1982, 75.
11. Thomas Pyzdek, "Non-Normal Distributions in the Real World," Quality Digest, December 1999, 36M1. 12. Johannes Ledolter, and Claude W. Burrill, Statis¬ tical Quality Control (New York: John Wiley & Sons, 1999). 13. N. Raghu Kackar, "Off-Line Quality Control, Parameter Design, and the Taguchi Method," Journal of Quality Technology 17, no. 4 (October 1985), 176-188. 14. Bruce D. Nordwall, "ITT Uses Process Control Methods to Increase Plant Productivity," Aviation Week & Space Technology, May 11,1987, 69-74. 15. Eric Wasiloff and Curtis Hargitt, "Using DOE to Determine AA Battery Life," Quality Progress, March 1999, 67-71. © 1999, American Society for Quality. Reprinted with permission. 16. Joseph J. Pignatiello, Jr., and John S. Ramberg, "The Top 10 Triumphs and Tragedies of Genichi Taguchi," presented at the 35th ASQC/ASA Fall Tech¬ nical Conference, Lexington, KY, 1991. 17. Douglas C. Montgomery, Design and Analysis of Experiments (New York: John Wiley & Sons, Inc., 1996); Charles Lipson and Narendra J. Sheth. Statistical Design and Analysis of Engineering Experiments (New York: McGraw-Hill Book Co., 1973). 18. Kalyan Kumar Chowdhury, E.V. Gigo, and R. Raghavan, "Quality Improvement Through Design of Experiments: A Case Study," Quality Engineering, 12, no. 3 (2000), 407-416. Copyright 2000 by Marcel Dekker, Inc.
564
19. Courtesy of Donald B. Splaun, Jr., Manager, Advanced Manufacturing Technology, GE-Fanuc, Inc. 20. Based on an anecdote in W. Edwards Deming, Out of the Crisis (Cambridge, MA: MIT Center for Advanced Engineering Study, 1986).
Part 3
Six Sigma and the Technical System
21. Adapted from Britz, et al. (see Note 4). 22. Adapted from Britz, et al. (see Note 4).
BIBLIOGFLAPHY Breyfogle, Forrest W. III. Implementing Six Sigma, 2nd ed,, New York: John Wiley & Sons, 2003. Chatfield, Christopher. Statistics for Technology: A Course in Applied Statistics, 3rd ed. New York: CRC Press, 1983. Deming, W. Edwards. The New Economics for Industry, Government, Education, 2nd ed. Cambridge, MA: MIT Press, 2000. -. Out of the Crisis. Cambridge, MA: MIT Press,
2000. Duncan, Acheson J. Quality Control and Industrial Statistics, 5th ed. Homewood, IL: Richard D. Irwin, 1986. Griffith, Gary. Quality Technician's Handbook, 5th ed. New York: Prentice Hall, 2002. Gunter, Bert. "Process Capability Studies Part I:
What Is a Process Capability Study?" Quality Progress 24, no. 2 (February 1991), 97-99. Pyzdek, Thomas, The Six Sigma Handbook. New York: McGraw-Hill, 2003. Robbins, C. L., and W. A. Robbins. "What Nurse Managers Should Know about Sampling Techniques." Nursing Management 20, no. 6 (June 1989), 46M8. Scherkenbach, William W. Deming's Road to Con¬ tinual Improvement. Knoxville, TN: SPC Press, 1991. Scholtes, Peter R. "Communities as Systems," ASQC 50th Annual Quality Congress Proceedings, 1996, 258-265. Tedaldi, Michael, Fred Seaglione, and Vincent Russotti. A Beginner's Guide to Quality in Manufacturing. Milwaukee, WI: ASQC Quality Press, 1992.
Design for Six Sigma Tools for Concept Development
Quality Profiles: Dana Corporation-Spicer Driveshaft and 3M Dental Products Division Quality Function Deployment Concept Engineering Tools for Design Development
Design Failure Mode and Effects Analysis Reliability Prediction Tools for Design Optimization
Reliability Testing Measurement System Evaluation Process Capability Evaluation
QUALITY in Practice: Testing Audio Components at Shure, Inc. QUALITY IN Practice: Applying Quality Function Deployment to a University Support Service Review Questions Problems
The Taguchi Loss Function
Projects, Etc.
Optimizing Reliability
CASES
Tools for Design Verification
Hydraulic Lift Co. Bloomfield Tool Co.
During World War II, 60 percent of the aircraft destined for the Far East proved unserviceable; 50 percent of electronic devices failed while still in storage; the service life of electronic devices used in bombers was only 20 hours; and 70 percent of naval electronics devices failed.1 Clearly, our ability to improve product performance has increased immensely over the last half-century. Nevertheless, businesses and consumers are still plagued with product failures or service upsets. Computers occa¬ sionally arrive DO A—dead on arrival—and most readers have undoubtedly encoun¬ tered the "blue screen of death" that results from software crashes. Consumers don't receive the products they ordered, or are given inaccurate information. More serious problems have occurred, such as injuries or deaths resulting from automobile tire separation or defective infant cribs and toys. Most of these problems fundamentally result from poor design or inadequate design processes. Design for Six Sigma (DFSS) represents a set of tools and methodologies used in the product development process for ensuring that goods and services will meet
565
Part 3
566
Six Sigma and the Technical System
customer needs and achieve performance objectives, and that the processes used to make and deliver them achieve six sigma capability. DFSS consists of four principal activities:2 , 1. Concept development, in which product functionality is determined based upon customer requirements, technological capabilities, and economic realities 2. Design development, which focuses on product and process performance issues necessary to fulfill the product and service requirements in manufacturing or delivery 3. Design optimization, which seeks to minimize the impact of variation in produc¬ tion and use, creating a "robust" design 4. Design verification, which ensures that the capability of the production system meets the appropriate sigma level DFSS can be viewed as a key approach within the overall product development process that we discussed in Chapter 7. DFSS is relatively new, but it is rapidly becoming rec¬ ognized and incorporated into traditional product development processes, and is not only applied to engineered products, but also to business transactions and production processes. General Electric is one company that embraced DFSS. For example, in its 1998 annual report, GE stated that "Every new product and service in the future will be DFSS. . . . They were, in essence, designed by the customer, using all of the critical-to-quality performance features (CTQs) the customer wanted in the product and then subjecting these CTQs to the rigorous statistical Design for Six Sigma Process." In its 2000 annual report, GE noted that more than half its sales would come from Like Six Sigma itself, most tools for DFSS products in 2001. DFSS have been around for some hr this chapter, we focus on some of the time; its uniqueness lies in the more important tools and best practices that manner in which they are integrated support DFSS efforts. However, we barely into a formal methodology, driven scratch the surface of the full scope of DFSS, and by the Six Sigma philosophy, with clear business objectives in mind. we encourage you to refer to some of the more complete references at the end of this chapter.
TOOLS FOR CONCEPT DEVELOPMENT Concept development is the process of applying scientific, engineering, and business
knowledge to produce a basic functional design that meets both customer needs and manufacturing or service delivery requirements. Concept development is a highly creative activity that can be enhanced by such techniques as brainstorming and brainwriting—a written form of brainstorming—and is focused first on identifying potential ideas. After potential ideas have been identified, they are evaluated using cost/benefit analysis, risk analysis, and other techniques. Finally, the best concept is selected, often using some type of scoring matrix to weight the selection criteria. The first question one must ask during concept development is: What is the product (good or service) intended to do? In Chapter 4 we stressed the importance of understanding the voice of the customer; it is the starting point for concept develop¬ ment. How the voice of the customer is translated into physical or operational speci¬ fications and production processes for a product or service can mean the difference between a successful product and an outright failure. For example, consumers expect a camera to take good pictures. In developing a new camera, Japanese engineers at one company studied pictures developed at
Chapter 12
Design for Six Sigma
567
Quality Profiles
.media,
r
Dana Corporation—Spicer Driveshaft and 3M Dental Products Division Dana Corporation 3H Dental Products
Dana Corporation-Spicer Driveshaft Division (now Torque Traction Technologies, Inc.) is North America's largest independent manufacturer of driveshafts and related components for light, medium, heavy duty, and off-highway vehicles. Spicer Driveshaft has 17 manufacturing, as¬ sembly, and administrative facilities around the United States, employing more than 3,400 people. The company uses Customer Platform Teams as one of the focal points for identifying customer requirements and building and maintaining new business, product offerings, and customer relation¬ ships. These teams include sales, engineering, quality, and warranty personnel that use a variety of formal and informal methods to listen and learn from customers. All senior leaders are involved in a two-phase strategic planning process that ad¬ dresses long-term direction and short-term objec¬ tives, which are linked and aligned from head¬ quarters to the individual manufacturing plants. A comprehensive diversity plan is used to help develop candidates for promotion from within the organization, improve community involvement efforts, and establish a mentoring program. From 1997 to 1999, sales increased by nearly 10 percent; economic value added increased from $15 million to $35 million; inventory as a per¬ centage of sales decreased from 6.8 percent to 6.3 percent; and working capital decreased from 13 percent to 10.2 percent of sales. Internal defect rates decreased more than 75 percent from 1996 to 2000 and are approaching best-in-class levels. Employees are encouraged to develop and imple¬ ment changes and innovative ideas and evaluate their results. Ideas submitted by employees average about three per month, which is ap¬ proaching best-in-class. In 1999, almost 80 percent of ideas were implemented. The employee turn¬ over rate is below 1 percent, which is better than the best competitor; and the attendance rate has remained above 98 percent for the last six years. The division received a Baldrige Award in 2000. Launched in 1964, 3M Dental Products Divi¬ sion (DPD), a Baldrige winner in 1997, is a business
unit of 3M Corporation, manufacturing and mar¬ keting more than 1,300 products used by dentists around the world, including restorative and crown and bridge materials and dental adhesive and infection control products. Innovation is a key suc¬ cess factor enabling the company to be a leader in the competitive dental products marketplace. Insights into changing customer requirements— combined with knowledge of technological, soci¬ etal, and environmental trends—are the starting point for product and process innovations. Den¬ tists, distributors, and major suppliers are in¬ volved in the division's systematic approach to translating key customer requirements into design requirements, prototypes, and, ultimately, reliable, quality products. Examples of customer involve¬ ment include simulated operations on "Fletchers," mannequins with human-like mouth features and conditions. Dentists use these mannequins to eval¬ uate variations of prototype material or hardware products. Through this and other methods, cus¬ tomer feedback is received at least three times during the development cycle. The company's business performance management matrix pro¬ vides a systematic and comprehensive tool for aligning key business drivers and goals down, through, and across all business and functional emits. More than 40 cross-functional teams arrange new product introduction, solves problems, and manage and improve business processes. In 1997 new product sales accounted for 45 percent of total sales, accelerating from 12 percent in 1992. Additionally, the number of patents per employee, an indicator of innova¬ tion, is better than twice the rate of its closest competitor. 3M DPD has been the industry leader in overall satisfaction of its U.S. distribu¬ tors since 1989 and in overall satisfaction of den¬ tists since 1987.
Source: Malcolm Baldrige National Quality Award, Profiles of Winners, National Institute of Standards and Technology, Depart¬ ment of Commerce.
Part 3
568
Six Sigma and the Technical System
photo labs and talked with customers to determine the major causes of poor pictures. The three biggest problems were underexposures, out-of-focus, and out-of-film (attempting to take pictures past the end of the roll). They developed the first camera that included a built-in flash to prevent underexposure, an autofocus lens, and an automatic rewind feature. Today, most popular models have these features to meet customer requirements. Other design considerations include the product's weight, size, appearance, safety, life, serviceability, and maintainability. When decisions about these factors are dominated by engi¬ neering considerations rather than by customer Developing a basic functional design involves translating cus¬ requirements, poor designs that fail in the tomer requirements into measur¬ market are often the result. able technical requirements and, Technical requirements, sometimes called subsequently, into detailed design design characteristics, translate the voice of the specifications. customer into technical language, specifically into measures of product performance. For example, consumers might want portable stereos with "good sound quality." Tech¬ nical aspects of a stereo system that affect sound quality include the frequency response, flutter (the wavering in pitch), and the speed accuracy (inconsistency affects the pitch and tempo of the sound). Technical requirements are actionable; they lead to design specifications such as the electrical properties of stereo system compo¬ nents. Two of the most powerful tools for meeting technical requirements are quality function deployment and concept engineering. Quality Function Deployment
A major problem with the traditional product development process is that customers and engineers speak different languages. A customer might express a desire to own a car that is easy to start. The translation of this requirement into technical language might be "car will start within 10 seconds of continuous cranking." Or, a requirement that "soap leaves my skin feeling soft" demands translation into pH or hardness specifications for the bar of soap. The actual intended message can be lost in the translation and subsequent interpretation by design or production personnel. The Japanese developed an approach called quality function deployment (QFD) to meet customers' requirements throughout the design process and also in the design of production systems. The term, a translation of the Kanji characters used to describe the process, can sound confusing. QFD is a planning process to guide the design, manufacturing, and marketing of goods by integrating the voice of the cus¬ tomer throughout the organization. Through QFD, every design, manufacturing, and control decision is made to meet the expressed needs of customers. It uses a type of matrix diagram to present data and information. QFD originated in 1972 at Mitsubishi's Kobe shipyard site. Toyota began to develop the concept shortly thereafter, and has used it since 1977 with impressive results. Between January 1977 and October 1979, Toyota realized a 20 percent reduc¬ tion in start-up costs on the launch of a new van. By 1982, start-up costs had fallen 38 percent from the 1977 baseline, and by 1984, were reduced by 61 percent. In addition, development time fell by one-third at the same time quality improved. Xerox and Ford initiated the use of QFD in the United States in 1986 (at that time, more than 50 percent of major Japanese companies were already using the approach). Today, QFD is used successfully by manufacturers of electronics, appliances, clothing, and con¬ struction equipment, by firms such as General Motors, Ford, Mazda, Motorola, Xerox, Kodak, IBM, Procter & Gamble, Hewlett-Packard, and AT&T. The 1992 model
Chapter 12
Design for Six Sigma
Cadillac was planned and designed entirely with QFD. Two organizations, the Amer¬ ican Supplier Institute, Inc., a nonprofit organization, and GOAL/QPC, a Massachu¬ setts consulting firm, have publicized and developed the concept in the United States. At the strategic level, QFD presents a challenge and the opportunity for top man¬ agement to break out of its traditional narrow focus on results, which can only be mea¬ sured after the fact, and to view the broader process of how results are obtained. Under QFD, all operations of a company are driven by the voice of the customer, rather than by edicts of top management or the opinions or desires of design engineers. At the tac¬ tical and operational levels, QFD departs from the traditional product planning process in which product concepts are originated by design teams or research and develop¬ ment groups, tested and refined, produced, and marketed. Often, a considerable amount of wasted effort and time is spent redesigning products and production sys¬ tems until customer needs are met. If customer needs can be identified properly in the first place, then such wasteful effort is eliminated, which is the principal focus of QFD. Product objectives are better understood and interpreted during the production process QFD benefits companies through improved communication and team¬ because all key design information is captured work between all constituencies in and synthesized. This approach helps to the value chain, such as between understand trade-offs in design, and promote marketing and design, between consensus among managers. Use of QFD design and manufacturing, and focuses on the drivers of customer satisfaction between purchasing and suppliers. and dissatisfaction, making it a useful tool for competitive analysis of product quality by top management. Productivity as well as quality improvements generally follow QFD. Perhaps most significant, though, QFD reduces the time for new product develop¬ ment. QFD allows companies to simulate the effects of new design ideas and con¬ cepts. Through this benefit, companies can reduce product development time and bring new products into the market sooner, thus gaining competitive advantage. Details of the QFD process and its use are presented next. The House of Quality A set of matrixes is used to relate the voice of the customer to
a product's technical requirements, component requirements, process control plans, and manufacturing operations. The first matrix, the customer requirement planning matrix shown in Figure 12.1, provides the basis for the QFD concept. The figure demonstrates why this matrix is often called the House of Quality. Building the House of Quality consists of six basic steps: 1. 2. 3. 4. 5. 6.
Identify customer requirements. Identify technical requirements. Relate the customer requirements to the technical requirements. Conduct an evaluation of competing products or services. Evaluate technical requirements and develop targets. Determine which technical requirements to deploy in the remainder of the pro¬ duction/delivery process.
To illustrate the development of the House of Quality and the QFD process, the task of designing a new fitness center in a community with two other competing organizations is presented. Step 1: Identify customer requirements. The voice of the customer is the primary input to the QFD process. As discussed in Chapter 4, many methods can be used to gather valid customer information. The most critical and most difficult
569
Part 3
570
Six Sigma and the Technical System
Figure 12.1 The House of Quality
Interrelationships
Technical requirements
Voice of the customer
Relationships between customer requirements and technical requirements
Priorities of customer requirements
Competitive evaluation
Priorities of technical requirements
step of the process is to capture the essence of the customer's needs and expec¬ tations. The customer's own words are vitally important in preventing misin¬ terpretation by designers and engineers. Figure 12.2 shows the voice of the customer in the House of Quality for the fitness center, perhaps based on a tele¬ phone survey or focus groups. They are grouped into five categories: programs and activities, facilities, atmosphere, staff, and other. These groupings can easily be done using affinity diagrams, for example. Step 2: List the technical requirements that provide the foundation for the product or service design. Technical requirements are design characteristics that describe the customer requirements as expressed in the language of the designer or engi¬ neer. Essentially, they are the "hows" by which the company will respond to the "whats"—customer requirements. They must be measurable, because the output is controlled and compared to objective targets. For the fitness center, these requirements include the number and type of program offerings and equipment, times, staffing requirements, facility characteristics and mainte¬ nance, fee structure, and so on. Figure 12.3 adds this information to the House of Quality. The roof of the House of Quality shows the interrelationships between any pair of technical requirements. Various symbols denote these relationships. A typical scheme uses the symbol • to denote a very strong relationship, ° for a
Chapter 12
Design for Six Sigma
Figure 12.2 Voice of the Customer in the House of Quality
Has programs I want E
o O < cl
Programs are convenient Family activities available
(n CD
'o cc LL
Clean locker rooms Well-maintained equipment Safe place to be
0 0 JZ Q_ C/3
Equipment available when desired
E
* o c CD > c
Job/personnel scheduling
■
Accuracy
Personnel
Nonroutine situations
Secondary
Information handling
Primary
Documents handling
Customer quality criteria
Customer handling
^
Housekeeping
How? What?
Procedures
System capacity
Service facility facets
Resources (personnel)
a Weak
Layout
•Medium
Relative importance
■ Strong
Resources (equipment)
Planning
Reliability
Responsiveness
Assurance
■
■
■
■ ■
A
•
■
■ ■
■
■
A
■
• ■
• ■
■ A
■
• ■
•
•
•
A
■
■ A
■
•
1
Source: R. Natarajan et al., "Applying QFD to Internal Service System Design," Quality Progress, February 1999, 65-70. © 1999, American Society for Quality. Reprinted with permission.
of service. One recommendation was to incorporate a better document-handling procedure. The RRC took the documents (resumes, tests, papers, and so on) and placed them in one incoming work bin. Staff members were good at sorting through the work orders and almost always identified the highpriority orders and completed them in plenty of time. But occasionally a work order was not noticed until it was almost too late. These orders were rushed and tended to be more prone to errors. To prevent errors, procedures needed to be established that would separate the work orders into separate bins for high-priority jobs and normal jobs. Bins might also be placed at each
person's workstation so employees did not have to sort through everyone else's work. Thus, employ¬ ees would have a better idea of the work they had to do and might better allocate their time. As mentioned earlier, the RRC had no formal training program. Although it might not be fea¬ sible for such a small organization to have a formal training program, some training proce¬ dures had to be considered. Employees needed to be thoroughly trained on documenting and routing work orders. An error in the work order inevitably resulted in an error in the final docu¬ ment. For instance, if a worker was not familiar with the work orders, he or she might document
618
Part 3
margin size in the wrong location. The profes¬ sional, accustomed to seeing that information in a particular location, might not notice the correct margin settings, resulting in a flawed document. Another suggested improvement was the facility layout. Many self-serve machines and resources were scattered around the room (see Figure 12.24). If they were centrally located, a cus¬ tomer could immediately target his or her needs upon entering the facility. Also, the service counter had no identifiable queuing system. If customers met the service counter head-on when entering the facility and key personnel were strategically posi¬ tioned behind the counter, a quasi-service queue could be created as customers directed themselves to specific service personnel. A proposed new layout is suggested in Figure 12.26. This layout would also facilitate document flow with Jody's design. Candy's typing, and Marie's copying and folding, cutting, and binding in the same area. These recommendations resulted in changes that were within the control of RRC personnel and therefore considered feasible. For instance, the
Six Sigma and the Technical System
layout was changed to facilitate smoother traffic flow and improve customer contact with service providers. Self-serve equipment, such as copiers, computers, and printers were grouped in the new layout for easy access. Key Issues for Discussion
1. Do you agree with the relative importance of measures of the voice of the customer in Figure 12.25? Explain why these rankings are reasonable, or provide counterarguments for a different ranking. 2. Using the relative importance ratings of the customer attributes and setting a scale of 1 = weak, 3 = medium, and 5 = strong for the rela¬ tionship matrix, compute a weighted score for each of the technical requirements in Figure 12.25. Do your scores support the conclusions of the study in terms of the key service com¬ ponents to deploy in the QFD process? 3. What other recommendations might you suggest based on the information provided in this case?
Figure 12.26 RRC Proposed Layout Laser station
•
•
•
Jody
Candy
Marie
CD Q. O u
Student
CD CL O o
Student
High¬ speed duplicating machine
A
Paper folder/ cutter/ binder
Paper stock and materials
Service desk
•
Study table
Study table
Study table
Study table
Reference materials
Source: R. Natarajan et al., "Applying QFD to Internal Service System Design," Quality Progress, February 1999, 65-70. © 1999, American Society for Quality. Reprinted with permission.
Chapter 12
jjgj(
Design for Six Sigma
Review Questions
1. What is Design for Six Sigma? Explain the four basic elements of DFSS and the various tools and methodologies that comprise this body of knowledge. 2. What are the principal benefits of QFD? 3. Outline the process of building the House of Quality. What departments and functions within the company should be involved in each step of the process? 4. Explain concept engineering. Why is it an important tool for assuring quality in product and process design activities? 5. Explain the difference between nominal dimensions and tolerances. How should tolerances be realistically set? 6. What is design failure mode and effects analysis (DFMEA)? Provide a simple example illustrating the concept. 7. What is the importance of reliability and why has it become such a prominent area within the quality disciplines? 8. Define reliability. Explain the definition thoroughly. 9. What is the difference between a functional failure and a reliability failure? 10. What is the definition of failure rate? How is it measured? 11. Explain the differences and relationships between the cumulative failure rate curve and the failure rate curve. 12. How is the average failure rate over a time interval computed? 13. Explain the product life characteristics curve and how it can be used. 14. What is a reliability function? Discuss different ways of expressing this function. 15. Explain how to compute the reliability of series, parallel, and series-parallel sys¬ tems. 16. Describe different forms of product testing. 17. What does the term latent defect mean? 18. What is a "robust" design? 19. Explain the role of the Taguchi loss function in process and tolerance design. 20. Provide some examples of low-tech and high-tech measuring instruments (see Bonus Materials). 21. Describe the science of metrology. 22. What is the difference between accuracy, precision, and reproducibility? 23. What is calibration and why is it important to a good quality assurance system? 24. How is an R&R study performed? What is its purpose? 25. Explain the term process capability. How can process capability generally be improved? 26. What are the three major types of process capability studies? Describe the methodology of conducting a process capability study. 27. Define the process capability indexes, Cp, Cpl, and Cpu, and explain how they may be used to establish or improve quality policies in operating areas or with suppliers. 28. What are the advantages and disadvantages of the Cpm capability index (see Bonus Materials)?
620
Part 3
Six Sigma and the Technical System
Problems Note: Data sets for several problems in this chapter are available in the Excel work¬ book C12Data.xls on the CD-ROM accompanying this text. Click on the appropriate worksheet tab as noted in the problem (e.g., Prob. 12-1) to access the data. 1. Bob's Big Burgers conducted consumer surveys and focus groups and identi¬ fied the most important customer expectations as • Healthy food • Speedy service • An easy-to-read menu board • Accurate order filling • Perceived value Develop a set of technical requirements to incorporate into the design of a new facility and a House of Quality relationship matrix to assess how well your requirements address these expectations. Refine your design as necessary, based upon the initial assessment. 2. Bob's Big Burgers (Problem 1) acquired some additional information. It found that consumers placed the highest importance on healthy food, followed by value, followed by order accuracy and service. The menu board was only casu¬ ally noted as an important attribute in the surveys. Bob faces three major com¬ petitors in this market: Grabby's, Queenburger, and Sandy's. Studies of their products yielded the information shown in Table 12.5. Results of the consumer panel ratings for each of these competitors are shown in Table 12.6 (a 1-5 scale, with 5 being the best). Using this information, modify and extend your House of Quality from Problem 1 and develop a deployment plan for a new burger. On what attributes should the company focus its marketing efforts? 3. Fingerspring, Inc., is working on a design for a new personal digital assistant (PDA). Marketing staff conducted extensive surveys and focus groups with
Table 12.5 Competitors' Product Information Company
Price
Size (oz.)
Calories
Sodium (mg)
Fat (%)
Grabby's Queenburger Sandy's
1.55 2.25 1.75
5.5 7.5 6.0
440 640 540
75 95 80
13 23 16
Table 12.6 Consumer Panel Ratings Attribute
Grabby's
Queenburger
Sandy's
Menu board Order accuracy Healthy food Speedy service Taste appeal Visual appeal Value
4 4 4 3 2 3 5
4 5 2 5 4 4 3
5 3 3 4 3 3 4
Chapter 12
Design for Six Sigma
621
potential customers to determine the characteristics that the customers want and expect in a PDA. Fingerspring's studies have identified the most important customer expectations as • Initial cost • Reliability • Ease of use • Features • Operating cost • Compactness Develop a set of technical requirements to incorporate into the design of a Flouse of Quality relationship matrix to assess how well your requirements address these expectations. Refine your design as necessary based upon the initial assessment. 4. Firtgerspring, Inc. (Problem 3), faces three major competitors in this market: Harespring, Springbok, and Greenspring. It found that potential consumers placed the highest importance on reliability (measured by such things as freedom from operating system crashes and battery life), followed by compact¬ ness (weight/bulkiness), followed by flexibility (features, ease of use, and types of program modules available). The operating cost was only occasionally noted as an important attribute in the surveys. Studies of their products yielded the information shown in Table 12.7. Results of the consumer panel ratings for these competitors are shown in Table 12.8 (a 1-5 scale, with 5 being the best). Using this information, modify and extend your House of Quality from Problem 3 and develop a deployment plan for a new PDA. On what attributes should the com¬ pany focus its marketing efforts?
Table 12.7 Competitors' Product Information
Company
Price
Wt. (oz.)
Size (In.)
Features
Operating Program
Battery life (hrs)
Opr. Cpsts (Batt./Prg/)
Harespring Springbok Greenspring
575 195 450
4.0 7.5 8.8
4.8 x 3.2 5.1 x 3.3 5.3 x 3.3
15 9 12
PalmOS® Hardmark* Easyware**
50 12 25
High Low Moderately high
Note: PalmOS® is one of the most recognized operating software programs for PDAs. *New unproven software, unique to Springbok **Well-received proprietary software, used on many PDAs for several years
Table 12.8 Consumer Panel Ratings Attribute
Harespring
Springbok
Greenspring
Initial cost Reliability Ease of use Features Operating cost Weight Size
3 5 4 4 5 5 4
5 2 1 2 3 3 4
4 3 3 3 4 3 4
622
Part 3
Six Sigma and the Technical System
5. Given the cumulative failure curve in Figure 12.27, sketch the failure rate curve. 6. Compute the average failure rate during the intervals 0 to 30,30 to 60, and 60 to 90, and 0 to 100, based on the information in Figure 12.27. 7. The life of a watch battery is normally distributed with a mean of 1,000 days and standard deviation of 60 days. a. What fraction of batteries is expected to survive beyond 1,100 days? b. What fraction will survive fewer than 880 days? c. Sketch the reliability function. d. What length of warranty is needed so that no more than 10 percent of the bat¬ teries will be expected to fail during the warranty period? 8. Lifetred, Inc., makes automobile tires that have a mean life of 50,000 miles with a standard deviation of 3,000 miles. a. What fraction of tires is expected to survive beyond 56,000 miles? b. What fraction will survive fewer than 47,000 miles? c. Sketch the reliability function. d. What length of warranty is needed so that no more than 10 percent of the tires will be expected to fail during the warranty period? 9. Compute the failure rate for six transformers that were tested for 600 hours each, three of which failed after 100,175, and 350 hours. 10. Assuming an exponential distribution, a particular lightbulb has a failure rate of 0.002 units per hour. What is the probability of failure within 400 hours? What is the reliability function? 11. The MTBF of a circuit is 1,200 hours. Calculate the failure rate. 12. The MTBF for an Internet service provider's Web server unit is normally distributed with a mean of 180 days and a standard deviation of 10 days. Each failure costs the company $750,000 in lost computing time and repair costs. A shutdown for preventive maintenance can be scheduled dur¬ ing nonpeak times and will cost $500,000. As the manager in charge of computer operations, you are to determine whether a preventive maintenance program is worthwhile. What is your recommendation based on a 1% proba¬ bility of failure? A 0.5% probability of failure? Assume 365 operating days per year.
Figure 12.27 Cumulative Failure Curve
10
20
30
40
50 Hours
60
70
80
90
100
Chapter
12
Design for Six Sigma
623
13. For a particular piece of equipment, the probability of failure during a given week is as follows: _Week of Operation
1 2 3 4 5 6
Probability of Failure
0.25 0.08 0.07 0.10 0.20 0.30
Management is considering a preventive maintenance program that would be implemented at the end of a given week of production. The production loss and downtime costs associated with an equipment failure are estimated to be $2,500 per failure. If it costs $500 to perform the preventive maintenance, when should the firm implement the preventive maintenance program? What is the total main¬ tenance and failure cost associated with your recommendation, and how many failures can be expected each year? Assume 52 weeks of operation per year.
14. An electronic missile guidance system consists of the following components: Components A, B, C, and D have reliabilities of 0.96, 0.98, 0.90, and 0.99, respec¬ tively. What is the reliability of the entire system? 15. A manufacturer of portable radios purchases major electronic components as modules. The reliabilities of components differ by supplier. Suppose that the configuration of the major components is given by:
The components can be purchased from three different suppliers. The reliabilities of the components are as follows:
Component
Supplier 1
Supplier 2
Supplier 3
A B C
.95 .80 .90
.92 .86 .93
.94 .90 .85
Transportation and purchasing considerations require that only one supplier be chosen. Which one should be selected if the radio is to have the highest possible reliability? 16. In a complex manufacturing process, three operations are performed in series. Because of the nature of the process, machines frequently fall out of adjustment and must be repaired. To keep the system going, two identical machines are used at each stage; thus, if one fails, the other can be used while the first is repaired (see accompanying figure).
Part 3
624
Six Sigma and the Technical System
Production System
The reliabilities of the machines are as follows:
17.
18.
19.
20.
21.
22.
23.
Machine
Reliability
A B C
.80 .90 .98
a. Analyze the system reliability, assuming only one machine at each stage. b. How much is the reliability improved by having two machines at each stage? An automated production system consists of three operations: turning, milling, and grinding. Individual parts are transferred from one operation to the next by a robot. Hence, if one machine or the robot fails, the process stops. a. If the reliabilities of the robot, turning center, milling machine, and grinder are 0.98,0.94, 0.98, and 0.90, respectively, what is the reliability of the system? b. Suppose that two grinders are available and the system does not stop if one fails. What is the reliability of the system? Two scales were used to weigh the same 25 samples of hamburger patties for a fast-food restaurant in Australia. Results are shown in C12dataset file for Prob.12-18 on the student CD-ROM. The samples were weighed in grams, and the supplier has ensured that each patty weighs 114 grams. Which scale is more accurate? Which is more precise? Which is the better scale? A gauge repeatability and reproducibility study at Frankford Brake Systems collected the data found in the C12dataset file for Prob.12-19 on the student CDROM. Analyze these data. The part specification is 1.0 ± 0.06 mm. A gauge repeatability and reproducibility study was made at Precision Parts, Inc., using three operators, taking three trials each on identical parts. The data that can be found in the C12dataset file for Prob.12-20 on the student CD-ROM were collected. Do you see any problems after analyzing these data? What should be done? The part specification for a collar that was measured was 1.6 ± 0.2 inches. A genetic researcher is trying to test two laboratory thermometers (that can be read to 1/100,000th of a degree Celsius) for accuracy and precision. She mea¬ sured 25 samples with each and obtained the results found in the C12dataset file for Prob.12-21 on the student CD-ROM. The true temperature being measured is 0 degrees C. Which instrument is more accurate? Which is more precise? Which is the better instrument? A specification for the length of an auto part is 5.0 ± 0.05 centimeters (cm). It costs $25 to scrap a part that is outside the specifications. Determine the Taguchi loss function for this situation. A blueprint specification for the thickness of a dishwasher part is 0.300 ± 0.024 centimeters (cm). It costs $9 to scrap a part that is outside the specifications. Determine the Taguchi loss function for this situation.
Chapter 12
Design for Six Sigma
24. A team was formed to study the auto part described in Problem 22. While con¬ tinuing to work to find the root cause of scrap, the team found a way to reduce the scrap cost to $17.50 per part. a. Determine the Taguchi loss function for this situation. b. If the process deviation from target can be held at 0.020 cm, what is the Taguchi loss? 25. A team was formed to study the dishwasher part described in Problem 23. While continuing to work to find the root cause of scrap, they found a way to reduce the scrap cost to $4 per part. a. Determine the Taguchi loss function for this situation. b. If the process deviation from target can be held at 0.015 cm, what is the Taguchi loss? 26. Ruido Unlimited makes electronic soundboards for car stereos. Output voltage to a certain component on the board must be 10 ± 0.1 volts. Exceeding the limits results in an estimated loss of $50. Determine the Taguchi loss function. 27. An electronic component has a specification of 150 ± 4 ohms. Scrapping the component results in an $80 loss. a. What is the value of k in the Taguchi loss function? b. If the process is centered on the target specification with a standard deviation of 2 ohms, what is the expected loss per unit? 28. An automatic cookie machine must deposit a specified amount of 7.5 ± 0.1 grams (g) of dough for each cookie on a conveyor belt. If the machine either over- or underdeposits the mixture, it costs $0.04 to scrap the defective cookie. a. What is the value of k in the Taguchi loss function? b. If the process is centered on the target specification with a standard deviation of 0.05 g, what is the expected loss per unit? 29. A computer chip is designed so that the distance between two adjacent pins has a specification of 2.000 ± 0.002 millimeters (mm). The loss due to a defective chip is $4. A sample of 25 chips was drawn from the production process and the results, in mm, can be found in the CUdataset file for Prob.12-29 on the student CD-ROM. a. Compute the value of k in the Taguchi loss function. b. What is the expected loss from this process based on the sample data? 30. The average time to handle a call in a call processing center has a specification of 6 ±1.25 minutes. The loss due to a mishandled call is $12. A sample of 25 calls was drawn from the process and the results, in minutes, can be found in the C12dataset file for Prob.12-30 on the student CD-ROM. a. Compute the value of k in the Taguchi loss function. b. What is the expected loss from this process based on the sample data? 31. In the production of transformers, any output voltage that exceeds ± 25 volts is unacceptable to the customer. Exceeding these limits results in an estimated loss of $200. However, the manufacturer can adjust the voltage in the plant by changing a resistor that costs $1.75. a. Determine the Taguchi loss function. b. Suppose the nominal specification is 120 volts. At what tolerance should the transformer be manufactured? 32. In the transformer business mentioned in the previous problem, managers gathered data from a customer focus group and found that any output voltage that exceeds ±20 volts was unacceptable to the customer. Exceeding these limits
625
626
Part 3
33.
34.
35.
36.
37.
Six Sigma and the Technical System
results in an estimated loss of $250. However, the manufacturer can still adjust the voltage in the plant by changing a resistor that costs $1.75. a. Determine the Taguchi loss function. b. Suppose the nominal specification remains at 120 volts. At what tolerance should the transformer be manufactured? Two processes, P and Q, are used by a supplier to produce the same component, Z, which is a critical part in the engine of the Boring 778 airplane. The specifica¬ tion for Z calls for a dimension of 0.24 mm ± 0.03. The probabilities of achieving the dimensions for each process based on their inherent variability are shown in the table found in the Clldataset file for Prob.12-33 on the student CD-ROM. If k- 75,000, what is the expected loss for each process? Which would be the best process to use, based on minimizing the expected loss? A machining process has a required dimension on a part of 0.560 ± 0.015 inch. Twenty-five parts each were measured as found in the Clldataset file for Prob.12-34 on the student CD-ROM. What is its capability for producing within acceptable limits? Adjustments were made in the process discussed in Problem 34 and 25 more samples were taken. The results are given in the C12dataset file for Prob.12-35 on the student CD-ROM. What can you observe about the process? What is its capability for producing within acceptable limits now? From the data for Kermit Theatrical Products, construct a histogram and esti¬ mate the process capability. If the specifications are 24 ± 0.03, estimate the per¬ centage of parts that will be nonconforming. Finally, compute Cp, Cpu, and Cp/. Samples for three parts were taken as shown in the C12dataset file for Prob.12-36 on the student CD-ROM. Samples for three parts made at River City Parts Co. were taken as shown in the CUdataset file for Prob.12-37 on the student CD-ROM. Data set 1 is for part 1, data set 2 is for part 2, and data set 3 is for part 3. a. Calculate the mean and standard deviations for each part and compare them to the following specification limits:
Part
Nominal
Tolerance
1 2 3
1.750 2.000 1.250
±0.045 ± 0.060 ±0.030
b. Will the production process permit an acceptable fit of all parts into a slot with a specification of 5 ± 0.081 at least 99.73 percent of the time? 38. Omega Technology Ltd. (OTL) is a small manufacturing company that pro¬ duces various parts for tool manufacturers. One of OTL's production processes involves producing a Teflon® spacer plate that has a tolerance of 0.05 to 0.100 cm in thickness. On the recommendation of the quality assurance (QA) depart¬ ment and over objections of the plant manager, OTL just purchased some new equipment to make these parts. Recently, the production manager was receiving complaints from customers about high levels of nonconforming parts. He suspected the new equipment, but neither QA nor plant management would listen.
Chapter 12
Design for Six Sigma
The manager discussed the issue with one of his production supervisors who mentioned that she had just collected some process data for a study that the quality assurance department was undertaking. The manager decided that he would prove his point by showing that the new equipment was not capable of meeting the specifications. The data provided by the supervisor are shown in the C12 Dataset.xls file for Problem 12.38 on the student CD-ROM. Perform a process capability study on these data and interpret your results. 39. Suppose that a process with a normally distributed output has a mean of 55.0 and a variance of 4.0. a. If the specifications are 55.0 ± 4.00, compute Cp, Cpk, and Cpm. b. Suppose the mean shifts to 53.0 but the variance remains unchanged. Recom¬ pute and interpret these process capability indexes. c. If the variance can be reduced to 40 percent of its original value, how do the process capability indices change (using the original mean of 55.0)? 40. A process has upper and lower tolerance limits of 5.60 and 5.20, respectively. If the customer requires a demonstrated Cp of 2.0, what must the process capa¬ bility be? If both Cpu and Cp; must also be 2.0, determine the mean and standard deviation of the process, assuming a normal distribution of output. 41. Clearly demonstrate that Six Sigma requires Cp = 2.0 and Cpk = 1.5.
jjjSH Projects, Etc. 1. Using whatever "market research" techniques are appropriate, define a set of customer attributes for a. Purchasing books at your college bookstore b. A college registration process c. A hotel room used for business d. A hotel room used for family leisure vacations For each case, determine a set of technical requirements and construct the rela¬ tionship matrix for the House of Quality. 2. (This exercise would best be performed in a group.) Suppose that you were developing a small pizza restaurant with a dining area and local delivery. Develop a list of customer requirements and technical requirements and try to complete a House of Quality. What service standards might such an operation have? 3. Most children (and many adults) like to assemble and fly balsa-wood gliders. From your own experiences or from interviews with other students, define a set of customer requirements for a good glider. (Even better, buy one and test it to determine these requirements yourself.) If you were to design and manufacture such a product, how would you define a set of technical requirements for the design? Using your results, construct a relationship matrix for a House of Quality. 4. Fill in the following relationship matrix of a House of Quality for a screw¬ driver. By sampling your classmates, develop priorities for the customer attrib¬ utes and use these and the relationships to identify key technical requirements to deploy.
627
Six Sigma and the Technical System
Plastic handle
Rubber grip
Steel shaft
Interchangeable bits
Price
Part 3
Ratchet capability
628
Easy to use Does not rust Durable Comfortable Versatile Inexpensive Priority
5. Discuss and prepare a report with examples on how DFMEA might be used in a service application rather than in a pure product design application. 6. Conduct an R&R study with a team of your fellow students to measure a set of sharpened pencils of various sizes. Use both an ordinary ruler and a metric ruler? What conclusions do you reach? 7. Visit several of the following metrology Web sites or find some new ones and summarize new ideas, concepts, or findings that are not discussed in this chapter. http://www.kinematics.com http://www.sandia.gov/psl http:/ / www.metrology.org http://www.boulder.nist.gov http://www.nist.gov http://www.gecals.com 8. Visit a local machine shop, bakery, or similar factory to determine what type of measurements they perform, what instruments they use, how they use the data, and how they ensure the precision and accuracy of their instruments and gauges. Write a report of your findings.
Chapter 12
Design for Six Sigma
629
I
I I
I. Hydraulic Lift Co. The Hydraulic Lift Company (HLC) manufactures freight elevators and automotive lifts used in garages and service stations. Figure 12.28 shows a simplified diagram of a hydraulic lift. The check valve is an important component in the system. Its purpose is to control the flow of hydraulic oil from the oil reservoir to the cylinder when the elevator is rising. As the elevator descends, the rate at which oil flows from the cylinder back to the reser¬ voir is also controlled by the check valve. One of the most important parts of the check valve is the piston, which moves within the valve body as the valve is opened or closed. The quality manager at HLC noticed that scrap rates on the piston had been quite high over the past three years. Two models (part numbers 117227 and 117228) of check valve pistons are being manufac¬ tured. Because of extremely critical tolerances, these parts are among the most difficult ones pro¬ duced in the machine shop. A study to determine the magnitude of the problem revealed that approximately $2,200 per month worth of parts had been scrapped over the past three years (see Figure 12.29). This amount translates to about 14 percent of total production of the parts, a scrap rate that is considered unac¬ ceptable. About half of the defective items were
Figure 12.28 Simplified Diagram of Hydraulic Lift
scrapped due to inability of the process to hold a 0.4990/0.4985-inch tolerance on the valve stem (see Figure 12.30). The machining operation used to shape the valve stem is performed on a grinding machine, which should have the capa¬ bility of holding a tolerance within 0.001-0.002 inch under standard operating conditions. Manu¬ facturing engineers and the quality manager decided to do a process capability study on one part (no. 117227) to gather statistical data on the stem problem and make a recommendation for improvements. For the first step, an operator ran 100 parts using the standard production methods. Results of the study [see the histogram in Figure 12.31(a)] revealed that a machine problem existed. The data showed that a few parts were being produced out¬ side the specifications. In addition, the strange shape of the histogram for dimensions within the specification limits prompted an investigation into the possibility of instability of the process. The study team observed that the operator was con¬ stantly adjusting the machine setting to try to hold to the specified tolerance. As a check on machine capability, the team asked the operator to run 20 parts without adjusting the machine, which resulted in scrap-
630
Part 3
Six Sigma and the Technical System
Figure 12.29 Average Scrap Cost per Month 117227 & 117228 1300 1250 1200 1150 1100 1050 1000 950 900 850 800 750 700 650 600 550 500 450 400 350 300 250 200 150 100 50 0 0.4990 Dia.
■ ■
Contr.
Stem groove
I■II_ Bad seat
Slots
Reason for Scrap
Figure 12.30 Part No. 117227 Check Valve
Broken stem
2.125 Dia.
Length
Deburr
Chapter 12
Design for Six Sigma
631
Figure 12.31 Process Capability for Hydraulic Lift Company
0 corresponds to 0.49800"
ping six of 20 parts, a 30 percent scrap rate [see Figure 12.31(b)]. This test verified that the machine needed some major adjustments. The machine manufacturer was contacted and a technician was dispatched to the plant. A run of 30 parts was made to show how the machine operated. Twelve of the 30 pieces were defective, with the stem dimension out of tolerance [see Figure 12.31(c)]. The technician made the following adjustments: • Installed new gaskets • Cleaned machine, adding oil and coolant
• Loaded hand wheel bearing for more posi¬ tive control • Reset retard pressure on grind wheel • Adjusted stone dresser mechanism • Reset dwell time (time the grindstone stays on the workpiece after reaching final diameter) The results of these adjustments were signifi¬ cant. Another 30 parts were run, with only two falling outside the tolerance limits [Figure 12.31(d)], The team still did not consider the process to be fully satisfactory. The manufacturer's
632
Part 3
technician said that the grinder "ways" (channels on which the machine head travels) would have to be reground and that some parts in the machine would have to be replaced. This recommendation was made to management, who agreed to have the machine overhauled as required. After the work was completed, a run of 35 parts was made. The results, shown in Figure 12.31(e), showed that all parts were well within tolerance limits. As a final step, operators and
Six Sigma and the Technical System
maintenance personnel were instructed on the proper use and care of the machine. Discussion Questions
1. Using the histograms in Figure 12.31, esti¬ mate the process capability indexes for each situation. 2. What lessons can be learned in terms of per¬ forming process capability studies and inter¬ preting the results?
II. Bloomfield Tool Co. Bloomfield Tool Co. (BTC) is a small manufac¬ turing company that produces precision tools to order. One of BTC's production processes involves producing a metal spacer plate that has a tolerance of 0.05 to 0.100 cm in thickness. Recently, the QA manager was receiving complaints from cus¬ tomers about high levels of nonconforming parts. He suspected problems in the gauges used to check outgoing parts were to blame, because they had not been sent out for calibration in some time. However, it was expensive to do, so he wanted to be sure that it was needed. He decided to do a gauge R&R test and selected two experienced inspectors to perform the test using 15 identical parts, whose dimensions had been verified. Table 12.9 shows data provided by the supervisor. Thus, data were from two operators, two gauges, and 15 parts that were measured twice, independently, by the two operators.
A reliability engineer pointed out that varia¬ tions in the data could be traced to three causes: repeatability problems (equipment variation or EV), reproducibility variations (appraisal variation or AV), and process variation (part variation or PV). Thus, total variation (TV) could be seen as made up of repeatability and reproducibility (R&R) variation and PV. Team members had heard about EV, AV, and R&R in a training class on measure¬ ment, but realized that the last item was an obvious but important point that they had not pre¬ viously considered. They were given the following formulas for computation of the PV and TV: PV = Rp x K„ and TV = V(R&R)2 + (PV)2
Table 12.9 Data Set for R&R Study Sample 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
0.0650 0.1000 0.0850 0.0850 0.0550 0.0875 0.0920 0.0850 0.0890 0.0600 0.0680 0.0545 0.0800 0.0700 0.0750
Operator 1 0.0600 0.1050 0.0800 0.0950 0.0580 0.0915 0.0880 0.0800 0.0980 0.0700 0.0750 0.0500 0.0910 0.0805 0.0650
0.0650 0.1050 0.0820 0.0820 0.0485 0.0945 0.0990 0.0750 0.0920 0.0550 0.0670 0.0525 0.0870 0.0865 0.0650
Operator 2 0.0550 0.0950 0.0750 0.0940 0.0525 0.0925 0.0900 0.0700 0.0990 0.0640 0.0720 0.0495 0.0820 0.0830 0.0680
Chapter 12
Design for Six Sigma
633
The Rp value is obtained by calculating the range of the sample averages in a gauge study. The K3 value depends on the number of parts mea¬ sured in the study. Some of these values are found in Table 12.10. Discussion Questions
1. Calculate the R&R, process, and total varia¬ tion for the data. Using the TV as the divisor, calculate the percentage of total variation that the EV, AV, R&R, and PV encompass. (Note: These variations are not directly related to one another, so the percentages
will not total to 100 percent.) What conclu¬ sions can you draw about the variations that were observed? 2. Based on your analysis, what recommenda¬ tions could you make on how the measure¬ ment system could be improved? 3. What would you tell the production manager? Note: Readers who have need for profes¬ sional software for performing extensive R&R studies or keeping track of gauge records and calibration may wish to look at R&Rpack and GAGEpack software devel¬ oped and distributed by PQ Systems, Inc., PO Box 10, Dayton, OH 45475-0010.
Table 12.10 Numbers of Parts (n) and K3 Factors
n K3
5 2.08
6 1.93
7 1.82
8 1.74
9 1.67
10 1.62
11 1.57
12 1.54
13 1.51
14 1.48
15 1.45
ENDNOTES 1. Reliability Guidebook, The Japanese Standards As¬ sociation (Tokyo: Asian Productivity Organization, 1972), 4. 2. C. M. Creveling, J. L. Slutsky, and D. Antis, Jr., Design for Six Sigma in Technology and Product Develop¬ ment (Upper Saddle River, NJ, Prentice Hall, 2003). 3. L. P. Sullivan, "Quality Function Deployment: The Latent Potential of Phases III and IV," in A. Richard Shores (ed.), A TQM Approach to Achieving Manufacturing Excellence (Milwaukee, WI: ASQC Quality Press, 1990), 265-279. 4. Christina Hepner Brodie, "A Polaroid Notebook: Concept Engineering," Center for Quality of Management Journal 3, no. 2 (1994), 7-14. 5. Laura Horton and David Boger, "How Bose Cor¬ poration Applied Concept Engineering to a Service," Center for Quality of Management Journal 3, no. 2 (1994), 52-59. 6. Jennifer Reese, "Starbucks: Inside the Coffee Cult," Fortune, December 9, 1996,190-200. 7. Susan Dillingham, "A Little Gross Stuff in Food Is OK by FDA," Insight, May 22,1989, 25. 8. Alan Vonderhaar, "Audi's TT Coupe's Ever So Close," Cincinnati Enquirer, November 27,1999, FI, F2. 9. Statement made by Belinda Collins before the
House Subcommittee on Technology, Committee on Sci¬ ence, June 29,1995. 10. ASQC Automotive Division Statistical Process Con¬ trol Manual (Milwaukee, WI: American Society for Quality Control, 1986). 11. This section is adopted from NIST Calibration Services, available at http://www.nist.gov. 12. Paul F. McCoy, "Using Performance Indexes to Monitor Production Processes," Quality Progress 24, no. 2 (February 1991), 49-55; see also Fred A. Spring, "The Cpm Index," Quality Progress 24, no. 2 (February 1991), 57-61. 13. Helmut Schneider, James Pruett, and Cliff Lagrange, "Uses of Process Capability Indices in the Supplier Certification Process," Quality Engineering 8, no. 2 (1995-1996), 225-235. 14. Mark L. Crossley, "Size Matters. How Good Is your Cpk, Really?" Quality Digest, May 2000, 71-72. 15. Appreciation is expressed to Christine Schyvinck, VP Operations, Shure, Inc., for providing this case (October 2000). 16. Adapted from R. Nat Natarajan, Ralph E. Martz, and Kyosuke Kurosaka, "Applying QFD to Internal Ser¬ vice System Design," Quality Progress, February 1999, 65-70. © 1999, American Society for Quality. Reprinted with permission.
634
Part 3
Six Sigma and the Technical System
BIBLIOGRAPHY Boser, Robert 15., and Cheryl L. Christ. "Whys, Whens, and Hows of Conducting a Process Capability Study." Presentation at the ASQC/ASA 35th Annual Fall Technical Conference, Lexington, Kentucky, 1991. Gunter, Bert. "Process Capability Studies Part I: What Is a Process Capability Study?"Quality Progress 24, no. 2 (February 1991), 97-99. Tomas, Sam. "Six Sigma: Motorola's Quest for Zero Defects." AP1CS, The Performance Advantage, July 1991, 36-41. -. "What Is Motorola's Six Sigma Product Quality?" American Production and Inventory Control
Society 1990 Conference Proceedings. Falls Church, VA: APICS; 27-31. Tedaldi, Michael, Fred Seaglione, and Vincent Russotti. A Beginner's Guide to Quality in Manufacturing. Mil¬ waukee, WI: ASQC Quality Press, 1992. Zubairi, Mazhar M. "Statistical Process Control Management Issues." 1985 IIE Fall Conference Proceed¬ ings. Reprinted in Mehran Sepehri (ed.). Quest for Quality: Managing the Total System. Norcross, GA: Indus¬ trial Engineering & Management Press, 1987.
Tools for Process Improvement Process Improvement Methodologies The Deming Cycle
Quality Profiles: Armstrong World Industries Building Products Operations and Xerox Business Services FADE Juran's Breakthrough Sequence Creative Problem Solving
Basic Tools for Process Improvement Flowcharts Run Charts and Control Charts
Poka-Yoke (Mistake-Proofing) Process Simulation
Engaging the Workforce in Process Improvement Skills for Team Leaders Skills for Team Members
QUALITY IN Practice: Process Improvement on the Free-Throw Line QUALITY IN Practice: Improving Patient Services at Middletown Regional Hospital Review Questions
Histograms
Discussion Questions Problems
Pareto Diagrams
Projects, Etc.
Cause-and-Effect Diagrams
CASES:
Check Sheets
Scatter Diagrams
Other Tools for Process Improvement
Readilunch Restaurant National Furniture Janson Medical Clinic
Kaizen Blitz
hen Cincinnati City Manager Valerie Lemmie started her job in April 2002, she asked building inspectors whether the city had a "one-stop shop" for building per¬ mits. They said, "Sure. You stop here once, you stop there once, and you stop there once." What she found out was that a permit stops 473 times on its way from the ini¬ tial application to the printer! After spending a week at City Hall and taking notes on every step of the process, a consultant hired to analyze the Department of Buildings and Inspections ended up with about 30 feet of flowcharts that depicted the building permit process. Although Ms. Lemmie conceded that improvement wouldn't be easy, an assistant noted that a lot of people wanted to know how they could do their jobs 635
Part 3
636
Six Sigma and the Technical System
better. "They know everything that's wrong with it probably more than anyone else. And more than anyone else, they need to be part of the solution."1 Improving such a complex process t^kes a lot of work, as companies such as Arm¬ strong and Xerox (see Quality Profiles) certainly understand. Having the right tools is important and can make the task considerably easier. We introduced basic concepts of process improvement in Chapter 7. However, the focus in that chapter was on the philosophy of improvement in the broader context of process management. In this chapter we focus on a variety of tools and techniques to help individuals and teams in process improvement projects. Most of these tools are included in basic Six Sigma Green Belt training.
PROCESS IMPROVEMENT METHODOLOGIES
Numerous methodologies for improvement have been proposed over the years. We already looked at one of them in the context of Six Sigma in Chapter 10: DMAIC. In this section we review some other popular approaches; most are simple variations of each other, but understanding them may provide new and unique insights into problem solving for process improvement. The Deming Cycle
A part of the kaizen philosophy (see Chapter 7), is the use of the Deming cycle to guide and motivate improvement activities. The Deming cycle is a simple method¬ ology for improvement that was strongly promoted by W. Edwards Deming. It was originally called the Shewhart cycle after its original founder, Walter Shewhart, but was renamed the Deming cycle by the Japanese in 1950. The Deming cycle is com¬ posed of four stages: plan, do, study, and act (PDSA) as illustrated in Figure 13.1. (The third stage—study—was formerly called check, and the Deming cycle was known as the PDCA cycle. Deming made the change in 1990. "Study" is more appropriate; with only a "check," one might miss something. However, many people still use "check.")
Figure 13.1 The Deming Cycle
Chapter 13
Tools for Process Improvement
637
Quality Profiles Armstrong World Industries Building Products Operations Armstrong World Industries
and Xerox Business Services Armstrong World Industries Building Products Operations (BPO), headquartered in Lancaster, Pennsylvania, manufactures acoustical ceilings and wall panels and employs about 2,400 people, 85 percent of whom work at the operation's seven manufacturing plants in six states. All quality-focused changes, from redesigning jobs and operations to reorganizing its salesforce, are driven by thoroughly evaluated expectations of increases in customer value. More than half of the BPO workforce participates in the approxi¬ mately 250 improvement teams operating at any given time. The team objectives range from cor¬ recting specific operational problems at a plant to improving key business processes across the organization. All quality improvement teams are required to develop specific action plans and set goals that will have a measurable impact on one of the company's key business drivers: customer satisfaction, sales growth, operating profit, asset management, and high-performance organiza¬ tion (human resource capabilities). Change is purposeful, guided by information and evalua¬ tions pointing the way to improvements that will make a major difference in customer value, employee value, and shareholder value. BPO enhanced its information gathering and analyt¬ ical capabilities, and stepped up its bench¬ marking studies to better understand the dynamics of the market, competitors' perfor¬ mance, and its own business results. From these initiatives, scrap was cut by 38 percent, and man¬ ufacturing output per employee rose 39 percent between 1991 and 1995. From 1992 to 1994, notices of nonconformance sent to suppliers fell 32 percent and on-time delivery improved from 93 percent to 97.3 percent, while reducing the delivery window from 4 hours to 30 minutes. Xerox Business Services (XBS) provides doc¬ ument outsourcing services and consulting, in¬ cluding on-site management of mailrooms and
print shops and the creation, production, and management of documents. XBS puts heavy emphasis on benchmarking throughout its orga¬ nization, gathering comparative information from companies that are large, small, tenured, new, inside, or outside their industry. XBS uses a process called Managing for Results, an inte¬ grated planning and management process that cascades action plans into measurable objectives for each manager, supervisor, and front-line associate. The entire process, the company says, is designed to "align goals from the customer's line of sight to the empowered employee and throughout the entire organization." Key perfor¬ mance measures are deployed to all teams and individuals in the company, and the process enables them to design work systems more rapidly and with greater flexibility. XBS also develops strategic initiatives based on its under¬ standing of the division's strengths and weak¬ nesses as well as its reading of opportunities and threats. This analysis draws on the division's extensive competitive intelligence, "voice of the customer," and "voice of the market" informa¬ tion systems. In monthly and quarterly reviews, the effectiveness of work processes is assessed against performance measures. The division invests more than $10 million annually for training, and it is continually searching for in¬ novative learning approaches. Examples are mini-camps—designed to help employees contemplate and prepare for future changes in the way they work and in how XBS addresses evolving customer requirements—and each employee's personal learning plan that is regu¬ larly reviewed by assigned "coaches."
Source: Malcolm Baldrige National Quality Award, Profiles of Winners, National Institute of Standards and Technology, Depart¬ ment of Commerce.
Part 3
638
Six Sigma and the Technical System
The plan stage consists of studying the current situation and describing the process: its inputs, outputs, customers, and suppliers; understanding customer expectations; gathering data; identifying problems; testing theories of causes; and developing solutions and action plans. In the do stage, the plan is implemented on a trial basis, for example, in a laboratory, pilot production process, or with a small group of customers, to evaluate a proposed solution and provide objective data. Data from the experiment are collected and documented. The study stage determines whether the trial plan is working correctly by evalu¬ ating the results, recording the learning, and determining whether any further issues or opportunities need be addressed. Often, the first solution must be modified or scrapped. New solutions are proposed and eval¬ uated by returning to the do stage. In the last The Deming cycle focuses on both stage, act, the improvements become standardshort-term continuous improveized and the final plan is implemented as a ment and long-term organizational “current best practice" and communicated learning. throughout the organization. This process then leads back to the plan stage for identification of other improvement opportunities. Table 13.1 summarizes the steps in the Deming Cycle in more detail. The funda¬ mental premise is that improvement comes from the application of knowledge.2 This knowledge may be knowledge of engineering, management, or how a process oper¬ ates that can make a job easier, more accurate, faster, less costly, safer, or better meet customer needs. Three fundamental questions to consider are: • What are we trying to accomplish? • What changes can we make that will result in improvement? • How will we know that a change is an improvement? Through a process of learning, knowledge is developed. With this philosophy, one can easily see why the Deming cycle has been an essential element of Japanese quality improvement programs. The following example demonstrates how the Deming cycle can be applied in practice. The co-owners of a diner decided to do something about the long lines that occurred every day in their place of business.3 After discussions with their employees, several important facts came to light: • • • •
Customers waited in line for up to 15 minutes. Usually, tables were available. Many of their customers were regulars. People taking orders and preparing food were getting in each other's way.
To measure the improvement that might result from any change they made, they decided to collect data on the number of customers in line, the number of empty tables, and the time until a customer received the food ordered. In the plan stage, the owners wanted to test a few changes. They decided on three changes: 1. Provide a way for customers to fax their orders in ahead of time (rent a fax machine for one month). 2. Construct a preparation table in the kitchen with ample room for fax orders. 3. Devote one of their two cash registers to handling fax orders. Both the length of the line and the number of empty tables were measured every 15 minutes during the lunch hour by one of the owners. In addition, when the
Chapter 13
Tools for Process Improvement
639
Table 13.1 Detailed Steps in the Deming Cycle
Plan 1.
Define the process: its start, end, and what it does.
2.
Describe the process: list the key tasks performed and sequence of steps, people involved, equipment used, environmental conditions, work methods, and materials used.
3.
Describe the players: external and internal customers and suppliers, and process operators.
4.
Define customer expectations: what the customer wants, when, and where, for both external and internal customers.
5.
Determine what historical data are available on process performance, or what data need to be collected to better understand the process.
6.
Describe the perceived problems associated with the process; for instance, failure to meet customer expectations, excessive variation, long cycle times, and so on.
7.
Identify the primary causes of the problems and their impacts on process performance.
8.
Develop potential changes or solutions to the process, and evaluate how these changes or solutions will address the primary causes.
9. Select the most promising solution(s). Do 1. Conduct a pilot study or experiment to test the impact of the potential solution(s). 2.
Identify measures to understand how any changes or solutions are successful in addressing the perceived problems.
Study 1.
Examine the results of the pilot study or experiment.
2.
Determine whether process performance has improved.
3.
Identify further experimentation that may be necessary.
Act 1.
Select the best change or solution.
2.
Develop an implementation plan: what needs to be done, who should be involved, and when the plan should be accomplished.
3.
Standardize the solution, for example, by writing new standard operating procedures.
4.
Establish a process to monitor and control process performance.
Source: Adapted from Small Business Guidebook to Quality Management, Office of the Secretary of Defense, Quality Management Office, Washington, DC (1998).
15-minute line check was done, the last person in line was noted, and the time until that person got served was measured. In the do phase, the owners observed the results of the three measures for three weeks. In the study phase, they detected several improvements. Time in line went down from 15 minutes to an average of 5 minutes. The line length was cut to a peak average of 12 people, and the number of empty tables decreased slightly. In the act phase, the owners held a meeting with all employees to discuss the results. They
Part 3
640
Six Sigma and the Technical System
decided to purchase the fax machine, prepare phone orders in the kitchen with the fax orders, and use both cash registers to handle walk-up and fax orders. FADE
A variation of the Deming cycle and DMAIC used by a number of organizations, including the U.S. Coast Guard, is known by the acronym FADE:/ochs, analyze, develop, and execute. In the Focus stage, a team selects the problem to be addressed and defines it, characterizing the current state of the process, why change is needed, what the desired result should be, and the benefits of achieving that result. In the analyze stage, the team works to describe the process in detail, determine what data and information are needed, and develop a list of root causes for the problem. The develop stage focuses on creating a solution and implementation plan along with documentation to explain and justify recommendations to management who must allocate the resources. Finally, in the execute stage, the solution is implemented and a monitoring plan is established. Juran’s Breakthrough Sequence
According to Juran, all breakthroughs follow a commonsense sequence of discovery, organization, diagnosis, corrective action, and control, which he formalized as the "breakthrough sequence," and which can be summarized as follows: • Proof of the need: Managers, especially top managers, need to be convinced that quality improvements are simply good economics. Through data collection efforts, information on poor quality, low productivity, or poor service can be translated into the language of money—the universal language of top manage¬ ment—to justify a request for resources to implement a quality improvement program. • Project identification: All breakthroughs are achieved project-by-project, and in no other way. By taking a project approach, management provides a forum for converting an atmosphere of defensiveness or blame into one of constructive action. Participation in a project increases the likelihood that the participant will act on the results. • Organization for breakthrough: Organization for improvement requires a clear responsibility for guiding the project. The responsibility for the project may be as broad as an entire division with formal committee structures or as narrow as a small group of workers at one production operation. These groups provide the definition and agreement as to the specific aims of the project, the authority to conduct experiments, and implementation strategies. The path from problem to solution consists of two journeys: one from symptom to cause (the diagnostic journey) and the other from cause to remedy (the remedial journey), which must be performed by different individuals with the appropriate skills. • Diagnostic journey: Diagnosticians skilled in data collection, statistics, and other problem-solving tools are needed at this stage. Some projects will require full¬ time, specialized experts (such as Six Sigma Black Belts) while the workforce can perform others. Management-controllable and operator-controllable problems require different methods of diagnosis and remedy. • Remedial journey: The remedial journey consists of several phases: choosing an alternative that optimizes total cost (similar to one of Deming's points), imple¬ menting remedial action, and dealing with resistance to change. • Holding the gains: This final step involves establishing the new standards and procedures, training the workforce, and instituting controls to make sure that the breakthrough does not die over time.
Chapter 13
Tools for Process Improvement
Many companies have followed Juran's program religiously. A Xerox plant in Mitcheldean, England, for example, cut quality losses by 30 percent to 40 percent and won a national prize in Britain in 1984 for quality improvement using the Juran system.4
Creative Problem Solving Solving quality problems often involves a high amount of creativity. In Japanese, the word creativity has a literal translation as "dangerous opportunity." In the Toyota production system, which has become the benchmark for world-class efficiency, a key concept is soikufn—creative thinking or inventive ideas, which means capital¬ izing on worker suggestions. The chairman of Toyota once observed, "One of the fea¬ tures of Japanese workers is that they use their brains as well as their hands. Our workers provide 1.5 million suggestions a year, and 95 percent of them are put to practical use. There is an almost tangible concern for improvement in the air at Toyota."5 An effective problem-solving process that can easily be adapted to quality improvement stems from creative problem-solving (CPS) concepts advocated by Osborn and by Parnes.6 This strategy consists of the following steps: • • • • • •
Understanding the "mess" Finding facts Identifying specific problems Generating ideas Developing solutions Implementation
Notice that the plan stage in the Deming cycle, for example, actually consists of the first five steps; the do, study, and act stages deal more with implementation. In Juran's program, the "diagnostic and remedial journeys" are essentially the same as this process. Thus, understanding these steps will help improve the application of other problem-solving models. For example, at Bethesda Hospitals of Cincinnati, Ohio, both the Juran and Deming approaches are integrated as shown in Figure 13.2. The left side of the figure incorporates the essential elements of Juran's diagnostic/reme¬ How one approaches problem solving dial journeys. Once a solution is proposed, the is not as critical as doing it in a Deming cycle is then used to evaluate the solu¬ systematic fashion, whether one uses the Deming cycle, FADE, tion's effectiveness prior to implementation. Juran's approach, CPS, or some Not every approach is appropriate for all orga¬ hybrid variation. nizations; one must be chosen or designed to fit the organization's culture and people.
BASIC TOOLS FOR PROCESS IMPROVEMENT Six Sigma created a renewed focus on process improvement. Among the many tools that comprise the Six Sigma toolbox are seven simple tools: flowcharts, check sheets, histograms, Pareto diagrams, cause-and-effect diagrams, scatter diagrams, and control charts. The Japanese called them the Seven QC (quality control) Tools, and they have been used for decades to support quality improvement problem-solving efforts. Table 13.2 shows the primary applications of each tool in Six Sigma and creative problem¬ solving processes. You can easily see how they also apply in the Deming cycle or Juran's approach. They are designed simply so that workers at all levels can use them easily. We will briefly review each tool to explain its role in quality improvement.
641
Part 3
642
Six Sigma and the Technical System
Figure 13.2 Bethesda Hospital Process Improvement Model
Source: Reprinted with permission of Bethesda Hospital, Inc., 619 Oak Street, Cincinnati, OH 45241.
Flowcharts
To clearly define a Six Sigma or any process improvement project, one must first understand the process that creates the outputs that internal or external customers receive. This understanding sets the foundation for identifying critical to quality (CTQ) issues, selecting measurements, and identifying root causes of problems, identifying A flowchart or process map iden¬ non-value-added steps, and reducing variation. tifies the sequence of activities or Flowcharts are best developed by having the the flow of materials and informa¬ tion in a process. Flowcharts help people involved in the process—employees, the people involved in the process supervisors, managers, and customers—con¬ understand it much better and struct the flowchart. A facilitator provides objec¬ more objectively by providing a pic¬ tivity in resolving conflicts. The facilitator can ture of the steps needed to accom¬ guide the discussion through questions such as plish a task. "What happens next?" "Who makes the deci¬ sion at this point?" and "What operation is
Chapter 13
Tools for Process Improvement
643
Table 13.2 Application of the Seven QC Tools in Six Sigma
Tool
DMAIC Application
CPS Application
Flowcharts
Define, Control
Mess-finding
Check sheets
Measure, Analyze
Fact-finding
Histograms
Measure, Analyze
Problem-finding
Cause-and-effect diagrams
Analyze
Idea-finding
Pareto diagrams
Analyze
Problem-finding
Scatter diagrams
Analyze, Improve
Solution-finding
Control charts
Control
Implementation
performed at this point?” Quite often, the group does not universally agree on the answers to these questions due to misconceptions about the process itself or a lack of awareness of the "big picture.” Flowcharts can easily be created using Microsoft Excel using the features found on the drawing toolbar.7 Flowcharts help all employees understand how they fit into a process and who are their suppliers and customers. This realization then leads to improved communi¬ cation among all parties. By participating in the development of a flowchart, workers feel a sense of ownership in the process, and hence become more willing to work on improving it. If flowcharts are used in training employees, more consistency will be achieved. Flowcharts also help to pinpoint places where quality-related measure¬ ments should be taken. Once a flowchart is constructed, it can be used to identify quality problems as well as areas for productivity improvement. Questions such as "How does this operation affect the customer?" "Can we improve or even eliminate this operation?" or "Should we control a critical quality characteristic at this point?" trigger the identification of opportunities. The AT&T customer-supplier model that we introduced in Chapter 4 provides a way of building a detailed process flowchart. Start with the outputs, or customer requirements, and move backward through the process to identify the key steps needed to produce each output; stop when the process reaches the supplier input stage. AT&T calls this technique backward chaining.8 AT&T suggests the following steps: 1. Begin with the process output and ask, "What is the last essential subprocess that produces the output of the process?" 2. For that subprocess, ask, "What input does it need to produce the process output?" For each input, test its value to ensure that it is required. 3. For each input, identify its source. In many cases, the input will be the output of the previous subprocess. In some cases, the input may come from external sup¬ pliers. 4. Continue backward, one subprocess at a time, until each input comes from an external supplier. This technique can be applied to each subprocess to create a more detailed process description. Once a flowchart is constructed, several fundamental questions can be asked to analyze the process:
Part 3
644
Six Sigma and the Technical System
• Are the steps in the process arranged in logical sequence? • Do all steps add value? Can some steps be eliminated and should others be added in order to improve quality or operational performance? Can some be combined? Should some be reordered? • Are capacities of each step in balance; that is, do bottlenecks exist for which cus¬ tomers will incur excessive waiting time? • What skills, equipment, and tools are required at each step of the process? Should some steps be automated? • At which points in the system might errors occur that would result in customer dissatisfaction, and how might these errors be corrected? • At which point or points should quality be measured? • Where interaction with the customer occurs, what procedures and guidelines should employees follow to present a positive image?
nnm
Process mapping and analysis is a powerful tool. Using process mapping as a basis for improvement. Motorola reduced manufacturing time for pagers from 40 days to less than one hour. Citibank adopted this approach and reduced internal call¬ backs in its Private Bank group by 80 percent and the credit process time by 50 per¬ cent. Its Global Equipment Finance division, which provides financing and leasing services to Citibank customers, lowered the credit decision cycle from three days to one. Copeland Companies, subsidiaries of Travelers Life & Annuity, reduced the cycle time of processing statements from 28 days to 15 days.9 The following example shows in more detail how Boise exploited process mapping.10 The Timber and Wood Products Division of Boise Cascade (now Boise) formed a team of 11 people with diverse backgrounds from manufacturing, administration, and marketing to improve a customer claims processing and tracking system that affected all areas and customers in its six divisions. Although external customer sur¬ veys indicated that the company was not doing badly, internal opinions of the oper¬ ation were far more critical. The first eye-opener came when the process was flowcharted and the group dis¬ covered that more than 70 steps were performed for each claim. Figure 13.3 shows the original flowchart from the marketing and sales department. Combined division tasks numbered in the hundreds for a single claim; the marketing and sales portion of the flowchart alone consisted of up to 20 separate tasks and seven decisions, which sometimes took months to complete. Most of these steps added no value to the set¬ tlement outcome. The flowchart accomplished much more than just plotting Boise's time and efforts; it also helped build team members' confidence in each other and foster mutual respect. When they saw how each member was able to chart his or her part of the process and state individual concerns, everyone's reason for being on the team was validated. The group eliminated 70 percent of the steps for small claims in the original flowchart, as shown in Figure 13.4. Run Charts and Control Charts
A run chart is a line graph in which data are plotted over time. The vertical axis rep¬ resents a measurement; the horizontal axis is the time scale. The daily newspaper usually includes several examples of run charts, such as the Dow Jones Industrial Average. They can be used to track such things as production volume, costs, and cus¬ tomer satisfaction indexes. The first step in constructing a run chart is to identify the measurement or indi¬ cator to be monitored. In some situations, one might measure the quality characteris¬ tics for each individual unit of process output. For low-volume processes, such as
Sales representative informs customer
Figure 13.3 Original Flowchart from the Marketing and Sales Department
First-level manager sends to next level of management
Chapter 13 Tools for Process Improvement
645
646
Part 3
Six Sigma and the Technical System
Figure 13.4 New Small Adjustment Request Form Process Flowchart for Marketing and Sales Department
chemical production or surgeries, this approach would be appropriate. However, for highvolume production processes or services with large numbers of customers or transactions, it would be impractical. Instead, samples taken on a periodic basis provide the data for computing basic statistical measures such as the mean, range or standard deviation, proportion of items that do not conform to specifications, or number of nonconformances per unit. Constructing the chart consists of the following
Run charts show the performance and the variation of a process or some quality or productivity indi¬ cator over time in a graphical fashion that is easy to understand and interpret. They also identify process changes and trends over time and show the effects of correc¬ tive actions. steps:
Step 1. Collect the data. If samples are chosen, compute the relevant statistic for each sample, such as the average or proportion. Step 2. Examine the range of the data. Scale the chart so that all data can be plotted on the vertical axis. Provide some additional room for new data as they are col¬ lected. Step 3. Plot the points on the chart and connect them. Use graph paper if the chart is constructed by hand; a spreadsheet program is preferable. Step 4. Compute the average of all plotted points and draw it as a horizontal line through the data. This line denoting the average is called the center line (CL) of the chart. If the plotted points fluctuate in a stable pattern around the center line, with no large spikes, trends, or shifts, they indicate that the process is apparently under con¬ trol. If unusual patterns exist, then the cause for lack of stability should be investi¬ gated and corrective action should be taken. Thus, run charts can identify messes caused by lack of control. A control chart is simply a run chart to which two horizontal lines, called control limits are added: the upper control limit (UCL) and lower control limit (LCL), as illustrated in Figure 13.5. Control charts were first proposed by Walter Shewhart at Bell Laboratories in the 1920s and were strongly advocated by Deming. Control limits are chosen statistically to provide a high probability (generally greater than 0.99) that points will fall between these limits if the process is in control. Control limits make it easier to interpret patterns in a run chart and draw conclusions about the state of control. These issues will be discussed in more detail in Chapter 14.
Chapter 13
Tools for Process Improvement
Figure 13.5 Structure of a Control Chart
Measurement
If sample values fall outside the control limits or if nonrandom patterns occur in the chart, then special causes may be affecting the process; the process is not stable. The process should be examined and corrective action taken as appropriate. If evalu¬ ation and correction are done in real time, then the chance of producing noncon¬ forming product is minimized. Thus, as a problem-solving tool, control charts allow operators to identify quality problems as they occur. Of course, control charts alone cannot determine the source of the problem. Operators, supervisors, and engineers may have to resort to other problem-solving tools to seek the root cause. Consider the following example.The Joint Commission Accreditation of Health Care Organizations (JCAHO) monitors and evaluates health care providers ac¬ cording to strict standards and guidelines. Improvement in the quality of care is a principal concern. Hospitals are required to identify and monitor important quality indicators that affect patient care and establish "thresholds for evaluation" (TFEs), which are levels at which special investigation of problems should occur. TFEs pro¬ vide a means of focusing attention on nonrandom errors (that is, special causes of variation). A logical way to set TFEs is through control charts. For instance, a hospital collects monthly data on the number of infections after surgeries. These data are shown in Table 13.3. Hospital administrators are concerned about whether the high percentages of infections (such as 1.76 percent in month 12) are caused by factors other than randomness. A control chart constructed from these data is shown in Figure 13.6. (Note that if the control limits are removed, it becomes a simple run chart.) The average percentage of infections is 55/7995 = 0.688 percent. Using formulas described in Chapter 14, the upper control limit is computed to be 2.35 percent. None of the data points fall above the upper control limit, indicating that the variation each month is due purely to chance and that the process is stable. To reduce the infection rate, management would have to attack the common causes in the process. The upper control limit would be a logical TFE to use, because any value beyond this limit is unlikely to occur by chance. Management can continue to use this chart to monitor future data.
647
648
Part 3
Six Sigma and the Technical System
Table 13.3 Monthly Data on Infections After Surgery Month
Surgeries
Infections
Percent
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
208 225 201 236 220 244 247 245 250 227 234 227 213 212 193 182 140 230 187 252 201 226 222 212 219 223 191 222 231 239 217 241 220 278 255 225
1 3 3 1 3 1 1 1 1 0 2 4 2 1 2 0 1 1 1 2 1 0 2 2 1 2 1 0 3 1 2 1 3 1 3 1
0.48 1.33 1.49 0.42 1.36 0.41 0.40 0.41 0.40 0.00 0.85 1.76 0.94 0.47 1.04 0.00 0.71 0.43 0.53 0.79 0.50 0.00 0.90 0.94 0.46 0.90 0.52 0.00 1.30 0.42 0.92 0.41 1.36 0.36 1.18 0.44
7,995
55
Check Sheets Check sheets are special types of Check sheets are simple tools for data collection. data collection forms in which the Nearly any kind of form may be used to collect results may he interpreted on the data. Data sheets use simple columnar or tab¬ form directly without additional ular forms to record data. However, to generate processing. useful information from raw data, further pro¬ cessing generally is necessary. In manufacturing, check sheets similar to Figure 13.7 are simple to use and easily interpreted by shop personnel. Including information such as specification limits
Chapter 13
Tools for Process Improvement
Figure 13.6 Control Chart for Surgery Infections
makes the number of nonconforming items easily observable and provides an imme¬ diate indication of the quality of the process. For example, in Figure 13.7 a significant proportion of dimensions are clearly out of specification, with a larger number on the high side than the low side. A second type of check sheet for defective items is illustrated in Figure 13.8, which shows the type of defect and a tally in a resin production plant. Such a check sheet can be extended to include a time dimension so that data can be monitored and analyzed over time, and trends and patterns, if any, can be detected. Figure 13.9 shows an example of a defect location check sheet. Kaoru Ishikawa relates how this check sheet was used to eliminate bubbles in laminated automobile windshield glass.11 The location and form of bubbles were indicated on the check sheet; most of the bubbles occurred on the right side. Upon investigation, workers discovered that the pressure applied in laminating was off balance—the right side was receiving less pressure. The machine was adjusted, and the formation of bubbles was eliminated almost completely. Histograms
Histograms provide clues about the characteristics of the parent popula¬ tion from which a sample is taken. Patterns that would be difficult to see in an ordinary table of numbers become apparent.
A histogram is a basic statistical tool that graphically shows the frequency or number of observations of a particular value or within a specified group. The check sheet in Figure 13.7, for example, was designed to provide the visual appeal of a histogram as the data are
649
650
Part 3
Six Sigma and the Technical System
Figure 13.7 Check Sheet for Data Collection
No.
(Continuous data use) Check Sheet Product name
Date
Usage
Factory name
Specification
Section name
No. of inspections
Data collector
Total number
Group name
Lot number
Remarks
Dimensions
1.5 1.6 1.7 1.8 1.9 2.0 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 3.0 3.1 3.2 40
c
c
35
03 O O a> CO
30
CO
o o CD
00
25
20
/
15
10
o
/
1
1
/ Total Frequency
1
/
/ 1
2
6
13
10
16
19
17
12
16
20
17
13
11 8
5
/ 6
2
1
Source: K. Ishikawa, Guide to Quality Control (Tokyo: Asian Productivity Organization, 1982), 31.
tallied. For these data, one can easily determine the proportion of observations that fell outside the specification limits. Some cautions should be heeded when interpreting histograms. First, the data should be representative of typical process conditions. If a new employee is now operating the equipment, or some aspect of the equipment, material, or method has changed, then new data should be collected. Second, the sample size should be large
Chapter 13
Tools for Process Improvement
651
Figure 13.8 Defective Item Check Sheet
Check Sheet Product:__
Date;
Manufacturing stage: final insp.
Section:
Type of defect: scar, incomplete, misshapen
Inspector's name: l_ot no
Factory:
Order no. Total no. inspected: 2530
Remarks: all items inspected
Type Surface scars Cracks Incomplete Misshapen Others
Check
Subtotal
mmmmmmn mmmmm mmmmmmmmm/u llll mm Grand total
Total rejects
mmmmmmmmmmmm mmmmmi
32 23 48 4 8 115
86
Source: K. Ishikawa, Guide to Quality Control (Tokyo: Asian Productivity Organization, 1982), 33.
enough to provide good conclusions; the larger, the better. Various guidelines exist, but a suggested minimum of at least 50 observations should be drawn. Finally, any conclusions drawn should be confirmed through further study and analysis. Pareto Diagrams
Joseph Juran popularized the Pareto principle in 1950 after observing that a high pro¬ portion of quality issues resulted from only a few causes. He named this technique after Vilfredo Pareto (1848-1923), an Italian economist who determined that 85 per¬ cent of the wealth in Milan was owned by only 15 percent of the people. For instance, in analyzing costs in a paper mill, Juran found that 61 percent of total quality costs were attributable to one category—"broke," which is paper mill terminology for paper so A Pareto distribution is one in defective that it is returned for reprocessing. In which the characteristics observed an analysis of 200 types of field failures of are ordered from largest frequency to smallest. A Pareto diagram is a automotive engines, only five accounted for histogram of the data from the one-third of all failures; the top 25 accounted largest frequency to the smallest. for two-thirds of the failures. In a textile mill, three of fifteen weavers were found to account
Part 3
652
Six Sigma and the Technical System
Figure 13.9 Defect Location Check Sheet
Source: K. Ishikawa, Guide to Quality Control (Tokyo: Asian Productivity Organization, 1982), 34.
for 74 percent of the defective cloth produced. Pareto analysis clearly separates the vital few from the trivial many and provides direction for selecting projects for improvement. Pareto analysis is often used to analyze data collected in check sheets. One may also draw a cumulative frequency curve on the histogram, as shown in Figure 13.10. Such a visual aid clearly shows the relative magnitude of defects and can be used to identify opportunities for improvement. The most costly or significant problems stand out. Pareto diagrams can also show the results of improvement programs over time. They are less intimidating to employees who are fearful of statistics.
_
a
eEMMDE©
N—*
Rotor Clip
A good example of Pareto analysis is found at Rotor Clip Company, Inc., of Som¬ erset, New Jersey, a major manufacturer of retaining rings and self-tightening hose clamps, and a believer in the use of simple quality improvement tools.12 One applica¬ tion involved the use of a Pareto diagram to study rising premium freight charges for shipping retaining rings. The study covered three months in order to collect enough data to draw conclusions. The Pareto diagram is shown in Figure 13.11. The results were startling. The most frequent cause of higher freight charges was customer requests. The decision was made to continue the study to identify which customers consistently expedited their shipments and to work closely with them to find ways of reducing costs. The second largest contributor was the lack of available machine time. Once a die was installed in a stamping press, it ran until it produced the max¬ imum number of parts (usually a million) before it was removed for routine mainte-
Chapter 13
Tools for Process Improvement
653
Figure 13.10 Pareto Diagram
100 97 90
70
“D
CD
42
28 20
7 3
Figure 13.11 Pareto Diagram of Customer Calls
improperly
3 CD D
Part 3
654
Six Sigma and the Technical System
nance. Although this policy resulted in efficient utilization of tooling, it tied up the press and ultimately accounted for rush shipments. The policy was revised to limit die runs to fill orders more efficiently. Pareto diagrams help analysts to progressively focus in on specific problems. Figure 13.12 shows one example. At each step, the Pareto diagram stratifies the data to more detailed levels (or it may require additional data collection), eventually iso¬ lating the most significant issues. Cause-and-Effect Diagrams
Variation in process output and other quality problems can occur for a variety of rea¬ sons, such as materials, machines, methods, people, and measurement. The goal of problem solving is to identify the causes of problems in order to correct them. The cause-and-effect diagram is an important tool in this task; it assists the generation of ideas for problem causes and, in turn, serves as a basis for solution finding. Kaoru Ishikawa introduced the cause-andeffect diagram in Japan, so it is also called an A cause-and-effect diagram is a Ishikawa diagram. Because of its structure, it is simple graphical method for pre¬ often called a fishbone diagram. Tire general struc¬ senting a chain of causes and effects ture of a cause-and-effect diagram is shown in and for sorting out causes and orga¬ nizing relationships between vari¬ Figure 13.13. At the end of the horizontal line, a ables. problem is listed. Each branch pointing into the main stem represents a possible cause. Branches pointing to the causes are contributors to those causes. The diagram identifies the most likely causes of a problem so that further data collection and analysis can be carried out. Cause-and-effect diagrams are constructed in a brainstorming type of atmos¬ phere. Everyone can get involved and feel they are an important part of the problem¬ solving process. Usually small groups drawn from operations or management work with a trained and experienced facilitator. The facilitator guides attention to discus¬ sion of the problem and its causes, not opinions. As a group technique, the cause-andeffect method requires significant interaction between group members. The facilitator who listens carefully to the participants can capture the important ideas. A group can often be more effective by thinking of the problem broadly and consid¬ ering environmental factors, political factors, employee issues, and even government policies, if appropriate. To illustrate a cause-and-effect diagram, a major hospital was concerned about the length of time required to get a patient from the emergency department to an inpatient bed. Significant delays appeared to be caused by beds not being available. A quality improvement team tackled this problem by developing a cause-and-effect diagram. They identified four major causes: environmental services, emergency department, medical/surgery unit, and admitting. Figure 13.14 shows the diagram with several potential causes in each category. It served as a basis for further investi¬ gations of contributing factors and data analysis to find the root cause of the problem. Scatter Diagrams Scatter diagrams are the graphical component of regression analysis. Even though
they do not provide rigorous statistical analysis, they often point to important rela¬ tionships between variables, such as the percentage of an ingredient in an alloy and the hardness of the alloy. Typically, the variables in question represent possible causes and effects obtained from Ishikawa diagrams. For example, if a manufacturer suspects that the percentage of an ingredient in an alloy is causing quality problems
Chapter 13
Tools for Process Improvement
655
Figure 13.12 Use of Pareto Diagrams for Progressive Analysis Defects in Parts per Million (PPM)
Percent
Percent What electrical component contributes the most defects? The K2 relay.
K2
U101
Q101
Q7
Q5
Others
Offset
Open
Spec
Noisy
Flux
Percent
Thermal
Source: Small Business Guidebook to Quality Management, Office of the Secretary of Defense, Quality Management Office, Washington, D.C.
Part 3
656
Six Sigma and the Technical System
Figure 13.13 General Structure of Cause-and-Effect Diagram
Figure 13.14 Cause-and-Effect Diagram for Hospital Emergency Admission Problem
Environmental Services
Emergency Department
Chapter 13
Tools for Process Improvement
in meeting hardness specifications, an employee group might collect data from sam¬ ples on the amount of ingredient and hardness and plot the data on a scatter diagram. Statistical correlation analysis is used to interpret scatter diagrams. Figure 13.15 shows three types of correlation. If the correlation is positive, an increase in variable x is related to an increase in variable y; if the correlation is negative, an increase in x is related to a decrease in y; and if the correlation is close to zero, the variables have no linear relationship. At Rotor Clip, which we highlighted earlier, the effect of advertising expenditures on the bottom line had been difficult to assess.13 Management wanted to learn whether the number of advertising dollars spent correlated with the number of new customers gained in a given year. Advertising dollars spent by quarter were plotted against the number of new customers added for the same period for three consecu¬ tive years (see Figure 13.16). The positive correlation showed that heavy advertising
Figure 13.15 Three Types of Correlation
y
y
y
x
No Correlation
Figure 13.16 Scatter Diagram of New Customers versus Advertising Dollars
Dollars Spent-
Part 3
658
Six Sigma and the Technical System
was related to new customers. The results were fairly consistent from year to year except for the second quarter of the third year, in which an outlier clearly stood out from the rest. Advertising checked the m'edia schedule and discovered that experi¬ mental image ads dominated that particular period. This discovery prompted the advertising department to eliminate image ads from its schedule.
OTHER TOOLS FOR PROCESS IMPROVEMENT
A variety of tools, developed and refined over the years, support process improve¬ ment efforts. In this section, we review some of the more popular ones. Kaizen Blitz
Blitz teams are generally comprised of employees from all areas involved in the process who under¬ stand it and can implement changes on the spot. Improvement is immediate, exciting, and satis¬ fying for all those involved in the process. Some examples of using kaizen blitz at Magnivision include the following:14
A kaizen blitz is an intense and rapid improvement process in which a team or a department throws all its resources into an improvement project over a short time period, as opposed to tradi¬ tional kaizen applications, which are performed on a part-time basis.
• The molded lens department ran two shifts per day, using 13 employees, and after
40 percent rework, yielded 1,300 pieces per day. The production line was imbal¬ anced and work piled up between stations, which added to quality problems as the work-in-process was often damaged. After a three-day blitz, the team reduced the production to one shift of six employees and a balanced line, reducing rework to 10 percent and increasing yield to 3,500 per day, saving more than $179,000. • In Retail Services, a blitz team investigated problems that continually plagued employees, and discovered that many were related to the software system. Some of the same customer information had to be entered in multiple screens, sometimes the system took a long time to process information, and sometimes it was difficult to find specific information quickly. Neither the programmers nor the engineers were aware of these problems. By getting everyone together, some solutions were easily determined. Estimated savings were $125,000. Poka-Yoke (Mistake-Proofing)
Human beings tend to make mistakes inadvertently.15 Typical mistakes in production are omitted processing, processing errors, setup errors, missing parts, wrong parts, and adjustment errors. Such errors can arise from the following factors: • • • • • • •
Forgetfulness due to lack of concentration Misunderstanding because of the lack of familiarity with a process or procedures Poor identification associated with lack of proper attention Lack of experience Absentmindedness Delays in judgment when a process is automated Equipment malfunctions
Blaming workers not only discourages them and lowers morale, but also does not solve the problem. The poka-yoke concept was developed and refined in the early 1960s by the late Shigeo Shingo, a Japanese manufacturing engineer who developed the Toyota produc-
Chapter 13
Tools for Process Improvement
659
tion system.16 Shingo visited a plant and observed that the plant was not using any type of measurement or statistical process control system for tracking defects. When asked why, the manager replied that they did not make any defects to track! His investigation led to the development of a mistake-proofing approach called Zero Quality Control, or ZQC. ZQC is driven by simple and inexpensive inspection processes, such as successive checking, in which operators inspect the work of the prior operation before continuing, and self-checking, in which operators assess the quality of their own work. Poka-yokes are designed to facilitate this process or remove the human element completely. Poka-yoke is focused on two aspects: (1) prediction, or recognizing that a defect is about to occur and providing a warning, and (2) detection, or recognizing that a defect has occurred and stopping the process. Many applications of poka-yoke are decep¬ tively simple, yet creative. Usually, they are inexpensive to implement. One of Shingo's first poka-yoke devices involved a process at the Yamada Electric plant in which workers assemble a switch having two push buttons supported by two springs.17 Occa¬ sionally, the worker would forget to insert a spring under each button, which led to a costly and embarrassing repair at the customer's facility. In the old method, the worker would take two springs out of a large parts box and then assemble the switch. To pre¬ vent this mistake, the worker was instructed first to place two springs in a small dish in front of the parts box, and then assemble the switch. If a spring remains in the dish, the operator knows immediately that an error has occurred. The solution was simple, cheap, and provided immediate feedback to the operator. Many other examples can be cited: Poka-yoke (POH-kah YOH-kay) is an approach for mistake-proofing processes using automatic devices or methods to avoid simple human error.
• Machines have limit switches connected to warning lights that tell the operator when parts are positioned improperly on the machine. • A device on a drill counts the number of holes drilled in a work piece; a buzzer sounds if the work piece is removed before the correct number of holes has been drilled. • Cassette covers are frequently scratched when the screwdriver slips out of the screw slot and slides against the plastic covers. The screw design shown in Figure 13.17 prevents the screwdriver from slipping. • A metal roller is used to laminate two surfaces bonded with hot melted glue. The glue tends to stick to the roller and causes defects in the laminate surface. An investigation showed that if the roller were dampened, the glue would not stick. A secondary roller was added to dampen the steel roller during the process, preventing the glue from sticking.
Figure 13.17 A Poka-Yoke Example of Screw Redesign
Old Design
New Design
660
Part 3
Six Sigma and the Technical System
• One production step at Motorola involves putting alphabetic characters on a keyboard, and then checking to make sure each key is placed correctly. A group of workers designed a clear template with the letters positioned slightly off center. By holding the template over the keyboard, assemblers can quickly spot mistakes. • Computer programs display a warning message if a file that has not been saved is to be closed. • A 3.5-inch diskette is designed so that it cannot be inserted unless the disk is ori¬ ented correctly (try it!). These disks are not perfectly square, and the beveled right corner of the disk allows a stop in the disk drive to be pushed away if it is inserted correctly. • Power lawn mowers now have a safety bar on the handle that must be engaged in order to start the engine. ® A proxy ballot for an investment fund will not fit into the return envelope unless a small strip is detached. The strip asks the respondent to check to see whether the ballot is signed and dated. From this discussion and examples, we see three levels of mistake-proofing with increasing costs associated with them: 1. Designing potential errors out of the product or process. Clearly, this approach is the most powerful form of mistake-proofing because it eliminates any possibility that the error or defect might occur and has no direct cost in terms of time or rework and scrap. 2. Identifying potential defects and stopping a process before the defect is produced. Although this approach eliminates any cost associated with producing a defect, it does require the time associated with stopping a process and taking corrective action. 3. Finding defects that enter or leave a process. This approach eliminates wasted resources that would add value to nonconforming work, but clearly results in scrap or rework. It is not always possible to achieve the highest level in designing or improving a process, but it is certainly advantageous to try. Richard B. Chase and Douglas M. Stewart suggest that the same concepts can be applied to services.18 The major differences are that service mistake-proofing must account for the customers' activities as well as those of the producer, and fail-safe methods must be set up for interactions conducted directly or by phone, mail, or other technologies, such as ATM. Chase and Stewart classify service poka-yokes by the type of error they are designed to prevent: server errors and customer errors. Server errors result from the task, treatment, or tangibles of the service. Customer errors occur during preparation, the service encounter, or during resolution. Task errors include doing work incorrectly, work not requested, work in the wrong order, or working too slowly. Some examples of poka-yoke devices for task errors are computer prompts, color-coded cash register keys, measuring tools such as McDonald's french-fry scoop, and signaling devices. Hospitals use trays for surgical instruments that have indentations for each instrument, preventing the surgeon from leaving one of them in the patient. Treatment errors arise in the contact between the server and the customer, such as lack of courteous behavior, and failure to acknowledge, listen, or react appropriately to the customer. A bank encourages eye contact by requiring tellers to record the cus¬ tomer's eye color on a checklist as they start the transaction. To promote friendliness
Chapter 13
Tools for Process Improvement
at a fast food restaurant, trainers provide the four specific cues for when to smile: when greeting the customer, when taking the order, when telling about the dessert special, and when giving the customer change. They encourage employees to observe whether the customer smiled back, a natural reinforcer for smiling. Tangible errors are those in physical elements of the service, such as unclean facil¬ ities, dirty uniforms, inappropriate temperature, and document errors. Hotels wrap paper strips around towels to help the housekeeping staff identify clean linen and show which ones should be replaced. Spell-checkers in word processing software help reduce document misspellings (provided they are used!). Customer errors in preparation include the failure to bring necessary materials to the encounter, to understand their role in the service transaction, and to engage the correct service. A computer manufacturer provides a flowchart to specify how to place a service call. By guiding the customers through three yes-or-no questions, the flowchart prompts them to have the necessary information before calling. Customer errors during an encounter can be due to inattention, misunderstanding, or simply a memory lapse, and include failure to remember steps in the process or to follow instructions. Poka-yoke examples include height bars at amusement rides that indicate rider size requirements, beepers that signal customers to remove cards from ATM machines, and locks on airplane lavatory doors that must be closed to turn on the lights. Some cashiers at restaurants fold back the top edge of credit card receipts, holding together the restaurant's copies while revealing the customer's copy. Customer errors at the resolution stage of a service encounter include failure to signal service inadequacies, to learn from experience, to adjust expectations, and to execute appropriate post-encounter actions. Hotels might enclose a small gift certificate to encourage guests to provide feedback. Strategically placed tray-return stands and trash receptacles remind customers to return trays in fast-food facilities. Mistake-proofing a service process requires identifying when and where failures generally occur. Once a failure is identified, the source must be found. The final step is to prevent the mistake from occurring through source inspection, self-inspection, or sequential checks. Process Simulation
Process simulation has been used routinely in business to address complex operational problems, so it is no wonder that it is a useful tool for Six Sigma applications, especially those involving customer service improve¬ ment, cycle time reduction, and reducing vari¬ Process simulation is an approach ability. Process simulation should be used to building a logical model of a real process, and experimenting with the when the process is highly complex and diffi¬ model to obtain insight about the cult to visualize, involves many decision points, behavior of the process or to evaluate or when the goal is to optimize the use of the impact of changes in assumptions resources for a process.19 or potential improvements to it. Building a process simulation model in¬ volves first describing how the process oper¬ ates, normally using a process map. The process map includes all process steps, including logical decisions that route materials or information to different locations. Second, all key inputs such as how long it takes to perform each step of the process and resources needed must be identified. Typically the activity times in a process are uncertain and described by probability distributions, which normally makes it diffi¬ cult to evaluate process performance and identify bottlenecks without simulation. The intent is for the model to duplicate the real process so that "what if?" questions
661
662
Part 3
Six Sigma and the Technical System
can be easily evaluated without having to make time-consuming or costly changes to the real process. Once the model is developed, the simulation process repeatedly samples from the probability distributions of the input variables to create a distribu¬ tion of potential outputs. As an example, a common customer support process is the help desk or call center process responsible for answering and addressing customers' questions and complaints.20 Typically, customer satisfaction ratings of the help desk are low. Although this process is common, it is difficult to analyze with conventional Six Sigma tools. The measure phase usually identifies "time to resolve an issue" and "quality of the issue resolution" as the two CTQs. When these factors are measured, performance is generally at less than a 1-sigma level, so significant improvement potential exists. Help desks are much too complex to analyze using basic Six Sigma tools. Most help desks have two or three levels of support. When a call comes in, it often waits in a queue. When a level 1 person is available, he or she takes the call. If this person cannot resolve the issue, the call is forwarded to level 2. If the level 2 representative cannot resolve the call, it is forwarded to engineering or a similar support group. Between each of these levels, the call may wait in several more queues, or the cus¬ tomer may be asked to wait for a call back. By developing a process simulation model, a Black Belt can validate the model against the real process by collecting data for model inputs, running the model, and statistically matching the results with data collected during the measure phase. Once the model is validated, analysis can begin. Most simula¬ tion packages provide operational output data for all the process steps, resource uti¬ lization data, and additional variables tracked throughout the process. When the data are collected, it becomes a straightforward task to analyze it statistically, identify bottlenecks, develop proposed solutions, and rerun the simulation to confirm the results. To provide a simple illustration, suppose that in a phone support center, in¬ coming calls arrive randomly with an average time between calls of about 5 minutes and a support representative evaluates the nature of each problem.21 Each call may take anywhere between 30 seconds and 4 minutes, although most can be handled in about 2 minutes. The representative is able to resolve 75 percent of the calls immedi¬ ately. However, 25 percent of the calls require that other support representatives do research and make a return call to the customer. The research, combined with the return call, requires on average 20 minutes, although this time may vary quite a bit, from as little as 5 minutes to more than 35 minutes. Figure 13.18 shows the process map for this situation, including the support representative resources. It is difficult to perform a process simulation, even for such a simple process, without some type of commercial simulation software. For this example, we used a package called ProcessModel22 which facilitates the simulation process by allowing the model to be built by simply "dragging and dropping" the process map symbols on the computer screen, entering the appropriate data input descriptions, and run¬ ning the model. As the model runs, ProcessModel provides a visual animation of the process, allowing the user to see the buildup of calls at each support stage to gain insight into the system performance. Standard output reports, such as the one shown in Figure 13.19, are generated automatically. By examining these results (see the circled entries in the figure), we see that support problems waited in the Return Call inQ activity an average of more than 496 minutes, and as many as 51 calls were waiting at any one time. Thus, this activity should be identified as a problem area suitable for process improvement efforts. In
Chapter 13
Tools for Process Improvement
Figure 13.18 Process Map for Help Desk Simulation Model
the RESOURCES section, we see that Support 1 was busy about half the time, while Support 2 was busy nearly 100 percent of the time. Any time human resource utiliza¬ tion is greater than 80 percent for extended periods, the system will most likely result in long waiting times and queue lengths, requiring more resources or changes in the assignment of resources. The simulation results suggest that better allocation of resources should improve performance. To reduce the customer waiting time we might add additional support representatives or cross-train and share the existing representatives. The simulation model can easily be modified to incorporate these changes and the impacts on the results can be evaluated. Clearly, trying to do this in the real process would be costly and disruptive, with no guarantee that it will work. Simulation is a rich and complex topic. Many good books exist about process simulation and we encourage you to explore some of the references given in the bib¬ liography.
ENGAGING THE WORKFORCE IN PROCESS IMPROVEMENT People are key to process improvement. Good people create innovative ideas and find and solve process problems. Expert team members perform the Deming cycle or the Six Sigma DMAIC process. People initiate and implement process improvement projects. It is not so much a question of what things people need to know as it is a question Compared to the technical tools for of what things they need to know how to do. gathering and analyzing data, the One team or one team member can make or "soft skills"—those that involve break an improvement project or a Six Sigma people—such as project manage¬ ment and team facilitation, are more initiative. People skills can be learned, but difficult to teach and learn. often take more time than is available for a single project; thus, they should be a routine part of every employee's educational program. Many of these issues, such as team leader roles, leadership development, team development, motivation, and job design, were discussed in Chapter 6. Some of the essential elements for effective process improvement from a people perspective are a shared vision and behavioral skills. A shared vision can unify a team and provide the motivation for successfully implementing the project. Developing one generally requires team discussions early on; unfortunately, inexperienced project
663
664
Part 3
Six Sigma and the Technical System
Figure 13.19 ProcessModel Simulation Results Scenario Replication Simulation Time
= = =
Normal Run 1 of 1 48 hr
ACTIVITIES
Activity Name
Scheduled Hours
Take Call inQ Take Call Perform Research inQ Perform Research Return Call inQ Return Call
Capacity
40 40 40 40 40 40
999
1 999 10 999 1
Total Entries 504 504 114 110 109 58
Average Minutes Per Entry
Average Contents
Maximum Contents
0.21 0.45 5.34 0.91 22.56 0.07
5
1
11 1
4
1.01 2.17 112.50
XI 9.92^\ ( 496.78
) V^3.oq^/
ACTIVITY STATES BY PERCENTAGE (Multiple Capacity)
% Activity Name
Scheduled Hours
Take Call inQ Perform Research inQ Perform Research Return Call inQ
% Empty
Partially Occupied
% Full
84.85 10.50 8.67 2.06
15.15 89.50 91.33 97.94
0.00 0.00 0.00 0.00
40 40 40 40
ACTIVITY STATES BY PERCENTAGE (Single Capacity) Activity Name Take Call Return Call
Scheduled Hours
% Operation
% Idle
% Waiting
% Blocked
40 40
45.62 7.25
54.38 92.75
0.00 0.00
0.00 0.00
Units
Scheduled Hours
Number of Times Used
Average Minutes Per Usage
% Util
1 1
48 48
504 168
2.17 14.08
Scheduled Hours
% In Use
% Idle
% Down
40 40
45.62 98.58
54.38 1.42
0.00 0.00
RESOURCES Resource Name Support 1 Support 2
RESOURCE STATES BY PERCENTAGE Resource Name Support 1 Support 2
ENTITY SUMMARY (Times in Scoreboard time units)
Entity Name Call HardCall
Average Qty Cycle Time Processed (Minutes) 398 58
4.19 596.99
Average VA Time (Minutes)
Average Cost
2.18 24.99
0.43 004
X45T52'' V98.58
i
Current Contents
1 51 0
% Util 0.02 45.62 0.53 9.13 2.26 7.25
Chapter 13
Tools for Process Improvement
665
Figure 13.19 ProcessModel Simulation Results (continued) VARIABLES
Variable Name Avg BVATime Entity Avg BVA Time Call Avg BVA Time HardCall
Total Changes
Average Minutes Per Change
Minimum Value
Maximum Value
Current Value
Average Value
0.00 6.10 39.04
0 0 0
0 0 0
0 0 0
0 0 0
1 391 59
j leaders frequently bypass these discussions in an effort to get the project underway. People who are technically oriented often neglect behavioral skills, thinking that such skills are unnecessary in order to solve technical problems. Behavioral skills require both knowledge and practice. Part of Deming's foundation for "profound knowl¬ edge" (Chapter 3) was the requirement to study, learn, and use psychology to improve quality. Skills for Team Leaders
As we discussed, team members often assume the role of project leaders and project managers and yet must defer to superior knowledge of other team members and take on roles as followers. In an insightful book on team-based project management, James Lewis observed that people skills needed by project managers could also be easily applied to team members.23 These skills include the following: • • • • • • •
Conflict management and resolution Team management Leadership skills Decision making Communication Negotiation Cross-cultural training
Conflict management involves dealing proactively with disagreements that may occur when two or more technical experts get together. Team management involves ensuring that project members remain focused on the goals, time frame, and costs of their part of the project. Leadership skills require that the project leader guide the work of the team, including team development, while managing upward to the pro¬ ject champion and outward to other project teams and team leaders. Decision making requires that good decisions be made in a timely fashion. Communication channels must be established and maintained throughout the course of the project. Negotia¬ tion is needed in order to secure the resources required for successful project com¬ pletion. Cross-cultural training may involve team members of other nationalities, or it may simply involve people from different functional areas with divergent points of view. In either case, it is extremely important for team members to be able to listen and learn about different perspectives on shared project goals from team and nonteam people who may have widely differing thoughts about issues under consideration.
666
Part 3
Six Sigma and the Technical System
Skills for Team Members
Perhaps the two areas of greatest importance in team functioning for process improve¬ ment project team members are meetings and shared decision making. Meetings are important because they consume considerable valuable time of team members. Shared decision making is important because most individuals in organizations have more practice in receiving direction from a supervisor, or making an individual deci¬ sion in their own workplace. Shared decisions are new territory for many individuals. Peter Scholtes provides some rules for effective meetings24: • • • • • •
Use agendas. Have a facilitator. Take minutes. Draft the next agenda. Evaluate the meeting. Adhere to the "100-mile" rule.
Scholtes suggests the use of detailed agendas that include topics, a sentence about the importance of each, who will present them, the estimated time for each topic, and the type of item, such as discussion, decision, or information topics. A facilitator can keep the discussion on time and on target, prevent anyone from dominating or being over¬ looked, and help bring the discussion to a close. A scribe who takes minutes can record subjects, decisions, and who will be responsible for actions taken. Drafting the next agenda at the end of the meeting serves to set a plan of action for going forward. Evaluating the meeting incorporates a continuous improvement step. Adhering to the "100-mile" rule requires a commitment to focus on the meeting so clearly that "no one should be called from the meeting unless it is so important that the disruption would occur even if the meeting was 100 miles away from the workplace."25 Decision-making techniques abound in quality improvement literature. One of the most powerful is called the nominal group technique (NGT), developed to pro¬ vide a way to prioritize and focus on important project objectives in the project defi¬ nition stage.26 One of the major advantages of the technique is that it balances the power of each individual involved in the decision process. Key steps in the process include the following: 1. Request that all participants (usually 5-10 persons) write or say which problem or issue they feel is most important. 2. Record all problems or issues. 3. Develop a master list of problems or issues. 4. Generate and distribute to each participant a form that numbers the problems or issues in no particular order. 5. Request that each participant rank the top five problems or issues by assigning five points to their most important perceived problem and one point to the least important of their top five. 6. Tally the results by adding the points for each problem or issue. 7. The problem or issue with the highest number is the most important one for the team as a whole. 8. Discuss the results and generate a final ranked list for process improvement action planning.27 This approach provides a more democratic way of making decisions and helps indi¬ viduals to feel that they have contributed to the process.
Chapter 13
Tools for Process Improvement
667
in Practice Improvement on the Free-Throw Line28 Timothy Clark observed that in basketball games, his son Andrew's free-throw percentage averaged between 45 and 50 percent. Andrew's process was simple: Go to the free throw line, bounce the ball four times, aim, and shoot. To confirm these obser¬ vations, Andrew shot five sets of 10 free throws with an average of 42 percent, showing little varia¬ tion among the five sets. Timothy developed a cause-and-effect diagram (Figure 13.20) to identify the principal causes. After analyzing the diagram and observing his son's process, he believed that the main causes were not standing in the same place on the free-throw line every time and having an inconsistent focal point. They developed a new process in which Andrew stood at the center of the line and focused on the middle of the front part of the rim. The new process resulted in a 36 percent improvement in practice (Figure 13.21). Toward the end of the 1994 season, he improved his average to 69 percent in the last three games.
During the 1995 season, Andrew averaged 60 percent. A control chart (Figure 13.22) showed that the process was quite stable. In the summer of 1995, Andrew attended a basketball camp where he was advised to change his shooting technique. This process reduced his shooting percentage during the 1996 season to 50 percent. Flowever, his father helped him to reinstall his old process, and his percentage returned to its former level, also improving his confidence. Key Issues for Discussion
1. How does this application conform to Deming's PDSA cycle? 2. Design a check sheet that might be useful to collect data for this analysis. How might Pareto diagrams be used to enhance the analysis?
Figure 13.20 Free-Throwing Cause-and-Effect Diagram
Materials
People
Measurement
Source: Adapted from Timothy Clark and Andrew Clark, "Continuous Improvement at the Free-Throw Line," Quality Progress, October 1997, 78-80. © 1997, American Society for Quality. Reprinted with permission.
668
Part 3
Six Sigma and the Technical System
Figure 13.21 Free-Throwing Shots Made Before and After Implementing the Improvement
Source: Adapted from Timothy Clark and Andrew Clark, "Continuous Improvement at the Free-Throw Line," Quality Progress, October 1997, 78-80. © 1997. American Society for Quality. Reprinted with permission.
Figure 13.22 Determining Whether the Free-Throw Process Is Stable
Practice Session
Source: Adapted from Timothy Clark and Andrew Clark, "Continuous Improvement at the Free-Throw Line," Quality Progress, October 1997, 78-80. © 1997. American Society for Quality. Reprinted with permission.
Chapter 13
Tools for Process Improvement
669
Quality in Practice Improving Patient Services at Middletown Regional Hospital29 Middletown Regional Hospital (MRH) is a licensed 310-bed acute-care hospital located in Middletown, Ohio, a southwestern Ohio city about 35 miles from Cincinnati. The parent, MRHS Corporation, includes a major hospital and 20 off-site locations. MRH employs approximately 1,700 people across a four-county area and provides all major medical services with the exception of open-heart surgery. CEO Douglas W. McNeill led MRH to become a quality-driven organization. This drive for con¬ tinuous improvement can be seen from the top management all the way down to the front-line employees. Dedication to quality is also evident in the mission, vision, and value statements. Every department at MRH is required to develop perfor¬ mance improvement goals and indicators annu¬ ally. All employees are trained annually in tools and approaches developed by the Quality Man¬ agement Department. Because of these efforts, the hospital received a number of quality awards and recognitions, including winning the first Codman Award, presented by the Joint Commission on Accreditation of Healthcare Organizations (JCAHO) for quality in a health care organization, and being named one of the 100 top hospitals in the United States. An important part of ensuring smooth-running daily operations at MRH is Maintenance and Envi¬ ronmental Services, whose director is Jim Faze. This division consists of two departments—the 20person Maintenance Department, which is respon¬ sible for the power plant, grounds, and the general maintenance of the facility for the main campus and the 20 off-site locations; and the Environ¬ mental Services (EVS) Department, which employs 56 people and is responsible for linen, waste management, and cleaning services for the 650,000 sqare foot main facility. To focus on improving services to its internal and external cus¬ tomers, EVS developed a customer service moni¬ toring system (CSMS) with three distinct parts: the Press Ganey Customer Survey, written customer comments that MRH receives from the survey, and a seven-step quality improvement process. Patients fill out the Press Ganey Customer Sat¬ isfaction Survey after receiving services. This
survey asks a full range of questions relating to all aspects of care the patient received while at any facility. The survey is divided into four major ser¬ vice sectors: Inpatient Services, Ambulatory Ser¬ vices, Outpatient Surgery, and Emergency Services. A direct question about the cleanliness of the hospital is asked for all sectors except Emer¬ gency Services. Results of the questionnaire are ranked against all hospitals nationwide in the database for each sector. The data are also sorted by geographic region and hospital size for addi¬ tional benchmarking results. Approximately 450 hospitals use this survey, enhancing its value as a benchmarking tool. MRH receives the report quarterly. The second portion of the CSMS consists of written customer comments from the survey. These comments are forwarded to the Guest Rela¬ tions Department and then to the specific depart¬ ments for action. MRH receives both positive and negative comments from this source. EVS formally praises employees with positive comments and retrains those who receive negative comments. The third element of the CSMS is improve¬ ment. Any time a major problem is encountered, MRH associates have been taught to apply a seven-step process, which was developed and implemented by an external management consul¬ tant group, and is taught and reinforced during the annual employee training sessions. The steps of this process are: 1. 2. 3. 4. 5.
Activate organizational awareness. Seek environmental transformation. Identify and define the process. Determine measurements. Collect data using statistical process control (SPC). 6. Analyze and make recommendations. 7. Remeasure to assess improvement. One problem EVS recognized was that "A sig¬ nificant gap exists between the current level of perceived customer satisfaction and the manage¬ ment goal, as measured by the Press Ganey Cus¬ tomer Satisfaction Survey. The objective is to find ways to eliminate the gap between the current
670 level of 69th percentile ranking and the manage¬ ment goal of 85th percentile." At the time of this study, MRH had been using the Press Ganey Customer Satisfaction Survey for eleven quarters. One question pertained to the cleanliness of the facility; thus, the key issue for EVS was "How clean was the facility?" Two fac¬ tors created difficulties in attacking this problem. First, MRH's management only monitored the per¬ centile ranking in the Press Ganey Survey for each division, but no one had the responsibility for completely analyzing all of the survey data. The second factor was the speed at which MRH received the data. MRH and the EVS department had no opportunity for service recovery because the patients were no longer at the facility when MRH received their complaints. The return rates on the customer surveys had remained fairly constant at 24 percent over the past three years. Return rates for the hospital's three ser¬ vice sectors were: 26 percent for Inpatient Services, 30 percent for Ambulatory Services, and 16 percent for Outpatient Surgery. The raw data for this ques¬ tion can be found in Table 1 in the Chapter 13 data¬ base on the student CD-ROM. Three key outcomes were measured: the percentile rank, the mean score analysis, and correlation coefficients. The percentile ranking was favored historically by upper manage¬ ment. It is simply the ranking against all the other hospital facilities involved in the nationwide survey. This information was tracked by Faze on a quarterly basis, charted, and forwarded to upper management. Previously, nothing was done at the departmental level to analyze the data and use it for improvement. An x-bar & R chart for the percentile rank for the department over the past 11 quarters can be seen in the database (in Table 1) and shows stable, consistent results. The mean and range scores had also stayed very consistent over the past 11 quarters. It is important to understand the relationship between the mean score and the percentile ranking when analyzing the Press Ganey data. One might expect that a consistent mean score would be correlated with the percentile ranking. The %Rank vs. Mean graph in the database shows a line chart for the relationship between the mean scores and the per¬ centile rank for the last 11 quarters. It is obvious that no direct correlation appears between these two sets of data. A higher mean score does not necessarily coincide with a high percentile rank.
Part 3
Six Sigma and the Technical System
A related measure from the Press Ganey Survey is the correlation coefficient, which mea¬ sures the relative importance of a specific question to the overall score. The higher a question's rela¬ tive correlation, the more likely that the overall satisfaction score for the survey will go up when this question's score goes up. Similarly, an item with a high correlation coefficient will bring down the overall satisfaction score if the individual ques¬ tion score goes down. Essentially, it is an indicator of the importance of the service in a given area to the customer. The correlations over time for EVS are shown in the x-bar & R chart for the correla¬ tion coefficients. Customer concerns show the rela¬ tively high degree of interest and concern relating to measures of cleanliness in a hospital setting. The chart shows a process that is stable. However, revealing information was discovered about corre¬ lation coefficient differences among the three ser¬ vice areas. The correlation coefficients for the Ambulatory Services and the Outpatient Surgery are much higher than for Inpatient Services. This result shows that the cleanliness question has less of an impact on the overall score of the inpatient section than in Outpatient Services. Over a two-year period, EVS teams tried three different approaches to improving the survey scores: implementing a computerized cleaning assignment system, tracking and addressing responses to comments received on customer sur¬ veys, and developing and deploying a daily room checklist to be filled out by housekeepers. The computerized cleaning assignment system helped improve the Press Ganey scores early in the process, but had more of an impact on the internal customers. The survey comments suggested that customers did not know what services to expect on a daily basis. MRH formed a team that developed tent cards similar to the ones used in the hotel industry. The tent card serves to let customers know what their room should look like when they arrive, what daily services they can expect, and a phone number to contact EVS if their expectations are not being met. The name of the housekeeper and encouragement to contact him/her about any cleaning issues is also on the card. This approach reduced the number of negative written comments MRH receives. EVS uses the daily room checklist to provide documentation on what services have been performed and to hold housekeepers account¬ able to get the work done right the first time.
Chapter 13
Tools for Process Improvement
Management and associates in EVS at MRH continue to search for ways to continually improve service levels in their quest to reach the elusive 85th percentile level that is management's long¬ term goal. Key Issues for Discussion
1. How do MRH's Total Quality Improvement Implementation System steps compare to the Deming Cycle and the DMAIC improvement steps?
671 2. What quality improvement tools have been used by EVS to address the problem? 3. What other insights can you get from analysis of the charts on the CD-ROM? 4. What have the three initiatives contributed in the efforts to solve the stated problem? 5. What else would you recommend in order to close the gap between perceived quality and the quality levels that EVS and MRH man¬ agement wants to provide?
Review Questions 1. Contrast the different process improvement methodologies described in this chapter. How are they alike, and how do they differ? 2. What is the Deming cycle? Explain the four steps. 3. Why do messes arise in organizations? 4. Describe the key issues that organizations face in the fact-finding phase of the CPS process. 5. Describe some techniques used to generate ideas. 6. What issues must be addressed in the solution-finding and implementation phases of the CPS process? 7. List and explain the Original Seven QC Tools. In what phases of the CPS process might each be most useful? 8. What types of questions might one ask to identify opportunities for improve¬ ment with a process flowchart? 9. Describe a control chart. How does it differ from a run chart? 10. Describe different types of check sheets that are useful in quality improvement. 11. Explain the difference between a histogram and a Pareto diagram. Do they apply to the same types of data? 12. Describe the structure of a cause-and-effect diagram. 13. How do scatter diagrams assist in finding solutions to quality problems? 14. Why do people make inadvertent mistakes? How does poka-yoke help prevent such mistakes? 15. What are the components of a process simulation? In what types of problem solving is it most useful? 16. Describe the types of errors that service poka-yokes are designed to prevent. 17. List and describe the tools needed for running an effective meeting. 18. What are the steps required to perform the nominal group technique (NGT)?
Discussion Questions 1. Table 13.4 shows the Pepsi-Cola Company's three-step method for customer¬ valued process improvement. Discuss its differences and similarities to the Deming cycle.
672
Part 3
Six Sigma and the Technical System
Table 13.4 Pepsi-Cola Process Improvement Methodology
Steps 1. Start with the customer
Actions a. Understand and prioritize customer needs b. Establish customer measures and success criteria c. Select a process with most impact on customer needs
2. Understand ourselves and plan improvements
a. Analyze the current process involving the performers in each step b. Design improved process c. Establish process measures
3. Do it
a. Pilot-test improved process b. Implement improved process c. Stabilize process d. Go to Step 1 (continuously improve)
Source: Courtesy of Pepsi-Cola Co. Reprinted with permission of The Forum Corporation.
2. What types of defects or errors might the following organizations measure and improve as part of a Six Sigma initiative? a. A metropolitan bus company b. A local department store c. An electric power company d. Walt Disney World or a regional amusement park, such as Paramount or Six Flags e. Your college or university 3. Discuss what would be the most appropriate tool to use to attack each of these quality issues: a. A copy machine suffers frequent paper jams and users are often confused as to how to fix the problem. b. The publication team for an engineering department wants to improve the accuracy of its user documentation but is unsure of why documents aren't error-free. c. An office manager has experienced numerous problems with a laser printer: double-spaced lines, garbled text, lost text, and blank pages. She is trying to figure out which is the most significant problem. d. A military agency wants to evaluate the weight of personnel at a certain facility. e. A bank needs to determine how many teller positions, drive-through sta¬ tions, and ATM machines it needs for a new branch bank in a certain busy location. Its information includes the average numbers and types of cus¬ tomers served by other similar facilities, as well as demographic information to suggest the level of customer traffic in the new facility.
Chapter 13
Tools for Process Improvement
f. A contracting agency wants to investigate why they had so many changes in their contracts. They believe that the number of changes may be related to the dollar value of the original contract or the days between the request for proposal and the contract award. g. A travel agency is interested in gaining a better understanding of how call volume varies by time of year in order to adjust staffing schedules. 4. Tire following two summaries of quality improvement projects performed by employee teams at Siemens Energy and Automation and Lucas Sumitomo Brakes, Inc., were presented in the 1997 Ohio Manufacturers' Association Case Studies in Team Excellence competition. Discuss how each example can be viewed in the context of (1) the Deming cycle, and (2) the creative problem solving process. Siemens Energy and Automation: Makin' Waves30 The Makin' Waves team is a Continuous Improvement team from the Siemens facility located in Urbana, Ohio. The Urbana facility is a supplier plant to the Siemens plant in Bellefontaine. We supply molded plastic, stamping, and plating support to the Bellefontaine plant. The Makin' Waves team is from the Plastics Department. Our team has been functioning for four years and has completed many highly successful projects. The team consists of two press operators, one product repairperson, one janitor, and one Quality Assurance person, all from the Plastics Department. We also included a supervisor from the E-Frame circuit breaker line in Bellefontaine who was added at the beginning of this project as a representative of our stakeholders and to provide valuable input. Our team began this project by looking into ideas for a project from the Corrective Action System and the Value Improvement Program. Our project started out as a way to reduce the negative effect caused by the poor appearance of the E-Frame breaker. Upon investigating the problem we dis¬ covered that we could actually eliminate the operation that was causing the negative appearance. We decided to make our project the elimination of the washing operation station in the production of the E-Frame plastic case. The E-Frame breaker case is molded in a compression press. The problem begins during the trimming and filing processes that are done after the part is removed from the mold. The plastic contains fiberglass, which becomes a fine dust that adheres to the part. To eliminate the dust, the parts are put through a washing operation. This process uses a conveyor system to carry the parts through a water spray cleaning system. The problem with this process is that the finish comes out looking spotty and with some fiberglass particles still adhering to the parts themselves. Our customers on the E-Frame breaker line had written Corrective Actions against this procedure because of the poor appearance and the dust still being present on the parts. They were experiencing problems with the fiberglass and had to wear gloves to protect their hands. Through data collection we realized that this operation takes 6,831 laborhours a year at a cost of over $96,000. Yet after the washing process the parts still were not clean and had a negative appearance that was not acceptable. We took the top five part numbers and charted the clean versus the dirty parts. We found that 97 percent of the parts did not meet customer standards and that our customers had to add a rework operation to keep the E-Frame line going! We set a goal to eliminate the washing operation by May 1997. In order to have this happen we needed to find a better process to take its place. We did a fishbone analysis to outline the causes of the problem, and followed up with a root cause analysis to eliminate any causes that did not pertain.
Part 3
674
Six Sigma and the Technical System
We brainstormed for possible solutions, producing five possible alterna¬ tives to the washing operation. They were: X
•
• • • •
*
Constant air flow Shop vacuum Deflashing parts Ionizer (mouse trap) Air hose at press
We tested and evaluated each solution, working with both the operators in the Plastics department and our customers on the breaker line. As a result of our evaluation, we found that an air hose at the press was the best solution. We instructed the operators that after the parts were filed, they should be blown free of all fiber particles. Because they were not being washed with water, this would eliminate the spotty appearance of the parts. We set up direct communication with our customers to make sure that this process was eliminating the problem permanently. Their feedback showed that they were satisfied with the new process and that there was not a problem with either the fiberglass or the appearance of the parts. We then took the findings and recommended that the washing operation be elimi¬ nated and replaced by an air hose at the press. We communicated to Quality Assurance that the job instructions should be updated to include our new process so that supervisors and operators would be trained on the new process at the end of their safety meetings. We then went to the sched¬ uler and had the washing process eliminated from the system. After this was accomplished and there was still favorable feedback from the cus¬ tomers, we pulled the plug on the washing operation altogether. Our goal as a team was to eliminate the washing operation and we accomplished this goal. There were other benefits attached to the project: • • • • •
• • •
$98,000 cost reduction in labor and maintenance Additional 136 square feet of valuable floor space freed up Improved delivery to customer Improved teamwork between customer and supplier Open communication with customer Elimination of a rework operation Improved quality to the consumer Improved safety and health of operators
Lucas Sumitomo Brakes, Inc.: Easy Money31 Lucas Sumitomo Brakes, Inc., has been manufacturing front disc brake calipers at its plant in Lebanon, Ohio, since 1989. As a result of this manu¬ facturing facility's success and worldwide reputation for quality, an expan¬ sion into manufacturing anti-lock braking system (ABS) components took place in early 1996. This expansion, and the subsequent start-up of produc¬ tion, created new challenges for the company and its employees. The company's basic philosophy is that employees are involved in the development and growth of the company and are encouraged to focus on the continuous improvement of processes to meet company goals. As soon as the ABS unit began production, some significant problem areas appeared, one of which was the large amount of downtime throughout the factory. In keeping with the company philosophy, the employees in the Maintenance Department formed a Continuous Improvement Team to address the downtime issue, and moved quickly to gather data. For a period of three months, each time a call for maintenance assistance was answered
Chapter 13
Tools for Process Improvement
within the factory, team members completed a Maintenance Response Report that indicated the machine associated with the downtime, the dura¬ tion of the downtime, and the root cause of the production delay. The infor¬ mation from these reports was entered into a computer database, and a Pareto chart was generated that indicated downtime by machine. This chart clarified that machines located on the Housing Machining Lines accounted for the majority of factory downtime; more specifically, two identical highpressure washing machines were causing approximately 80 percent of the Machining Line downtime. The team established a goal of reducing down¬ time associated with the high-pressure washers by 40 percent in the short term, and by 70 percent over the long term. Using the data collected, the team members brainstormed possible causes for the excessive washer downtime. The results pointed to three major recurring problems that, according to the data taken from the Mainte¬ nance Response Reports, accounted for approximately 99 percent of the downtime. These are: • • •
Proximity switch replacement and adjustments Repairing or replacing jig clamps or dryer clamps High pressure drops within the machine associated with the erosion of O-rings inside the rotary joints
Utilizing more brainstorming sessions, the fishbone diagram technique, and asking the five Whys (see Chapter TO), the team identified some pos¬ sible solutions to these recurring problems. As a result, it was able to pre¬ sent specific ideas for improvement to Engineering and Manufacturing, which enabled these departments to assist in communicating with the machine manufacturer, give the project the support that was needed to implement the countermeasures identified by the team, and allow for scheduling of Machine Line downtime to perform trials, which were essen¬ tial during a time when the Machine Lines were running production six days a week to meet customers' schedules in a just-in-time system. With this support, the team implemented the following corrective actions related to the problems associated with high-pressure washer downtime: •
•
•
The proximity switches and wires were relocated outside the washing machines to eliminate melted switches, and to prevent switches from being destroyed by the high-pressure water blast. High-temperature-resistant O-rings and seals were installed in the machines, and air tubing was changed to a larger size to eliminate machine-clamping problems. A preventive maintenance program was implemented to rebuild rotary joints during scheduled downtime to eliminate unscheduled downtime due to worn O-rings inside the rotary joints.
Following implementation of these improvements, the team members tracked downtime associated with the washing machines for a period of three months, and found that the results exceeded their expectations. A 45 percent decrease in downtime associated with the problems identified by the team was achieved, and average downtime per month was improved from 507 minutes to 276 minutes, saving more than a week of production time each year. Following these improvements, the team moved on to focus on other areas of high downtime within the factory, and washer downtime associated with the improvements made has continued to improve as pro¬ duction levels have increased.
675
676
Part 3
Six Sigma and the Technical System
5. Maintaining accuracy of books on the shelves in a college library is an impor¬ tant task. Consider the following problems that are often observed. a. Books are not placed in the correct shelf position, which includes those books that have been checked out and returned, as well as those taken off the shelves for use within the library by patrons. b. New or returned books are not checked in and consequently, the online cat¬ alog does not show their availability. What procedures or poka-yokes might you suggest for mitigating these prob¬ lems? You might wish to talk to some librarians or administrators at your col¬ lege library to see how they address such problems.
jjj^f Problems 1. A flowchart for a fast-food drive-through window is shown in Figure 13.23. Determine the important quality characteristics inherent in this process and suggest possible improvements. 2. A catalog order-filling process for personalized printed products can be described as follows:32 Telephone orders are taken over a 12-hour period each day. Orders are collected from each person at the end of the day and checked for errors by the supervisor of the phone department, usually the following morning. The super¬ visor does not send each one-day batch of orders to the data processing depart¬ ment until after 1:00 p.m. In the next step—data processing—orders are invoiced in the one-day batches. Then they are printed and matched back to the original orders. At this point, if the order is from a new customer, it is sent to the person who did the customer verification and setup of new customer accounts. This process must be completed before the order can be invoiced. The next step— order verification and proofreading—occurs after invoicing is completed. The orders, with invoices attached, are given to a person who verifies that all required information is present and correct to permit typesetting. If the verifier has any questions, they are checked by computer or by calling the customer. Finally, the completed orders are sent to the typesetting department of the print shop. a. Develop a flowchart for this process. b. Identify opportunities for improving the quality of service in this situation. 3. An independent outplacement service helps unemployed executives find jobs. One of the major activities of the service is preparing resumes. Three word processors work at the service typing resumes and cover letters. Together they handle about 120 individual clients. Turnaround time for typing is expected to be 24 hours. The word-processing operation begins with clients placing work in the assigned word processor's bin. When the word processor picks up the work (in batches), it is logged in using a time clock stamp, and the work is typed and printed. After the batch is completed, the word processor returns the docu¬ ments to the clients' bins, logs in the time delivered, and picks up new work. A supervisor tries to balance the workload for the three word processors. Lately, many of the clients have been complaining about errors in their documents— misspellings, missing lines, wrong formatting, and so on. The supervisor has told the word processors to be more careful, but the errors still persist. a. Develop a cause-and-effect diagram that might clarify the source of errors. b. What tools might the supervisor use to study ways to reduce the number of errors?
Chapter 13
Tools for Process Improvement
Figure 13.23 Flowchart for a Fast-Food Drive-Through Window (Problem 1)
4. The times required for trainees in an electronics course to assemble a compo¬ nent used in a computer were measured. These are shown in the C13dataset file for Prob. 13-4 on the student CD-ROM. Construct a histogram to graphically show the data. What recommendations for improvement would you give the course instructor, based on your findings? 5. A Six Sigma analyst in a bank suspected that errors in counting and manually strapping cash into bundles were related to the number of weeks that employees had been employed on that job. The data found in the C13dataset file for Prob. 13-5 on the student CD-ROM were gathered from the process. What do you conclude from your analysis? What do you recommend? 6. The data found in the C13dataset file for Prob. 13-6 on the student CD-ROM were gathered from a process used to make plastic gears for a computer printer. The gears were designed to be 2.5 ± 0.05 centimeters (cm) in diameter. Construct a histogram based on the data given. What can you observe about the shape of the distribution? What would you recommend to the production manager, based on your analysis? 7. The times required to prepare standard-size packages for shipping were mea¬ sured. These data are shown in the C13dataset file for Prob. 13-7 on the student CD-ROM. Construct a scatter diagram for these data. What recommendations for improvement would you give the section leader, based on your findings? 8. In a manufacturing process, the production rate (parts/hour) was thought to affect the number of defectives found during a subsequent inspection. To test this theory, the production rate was varied and the numbers of defects were col¬ lected for the same batch sizes. The results can be found in the C13dataset file for
677
Part 3
678
Six Sigma and the Technical System
Prob. 13-8 on the student CD-ROM. Construct a scatter diagram for these data. What conclusions can you reach? 9. Ace Printing Company realized that they were losing customers and orders due to various delays and errors. In order to get to the root cause of the problem, they decided to track problems that might be contributing to customer dissatis¬ faction. The following list of the problems found shows their frequencies of occurrence over a six-month period. What technique might you use to graphi¬ cally show the causes of customer dissatisfaction? What recommendations could you make to reduce errors and increase customer satisfaction? Error/Delay Cause
Frequency
Customer change delays Lack of press time Design department delays Paper not in stock Lack of proper order information Lost order Press setup delays
15 180 55 75 24 11 240
10. Rick Hensley owns an automotive dealership. Service is a major part of the operation. Rick and his service team spent considerable time in analyzing the service process and developed a flowchart, shown in Figure 13.24, that describes the typical activities in servicing a customer's automobile. Rick wants to ensure that customers receive superior service and are highly satisfied; thus, he wants to establish poka-yokes for any possible failures that may occur. Your assignment is to identify any possible failure in the service process that may be detrimental to customer satisfaction and suggest poka-yokes to eliminate these failures. 11. Figure 13.25 shows a medication administration process in a hospital. The administrative staff of the hospital is concerned about frequent medication errors. After examining this flowchart, discuss possible sources of errors, the types of individuals responsible (e.g., physicians, nurses, pharmacists, other), and poka-yokes that might be used to mitigate these errors. 12. Analysis of customer complaints for a large dot-com apparel house revealed the following: Billing errors Shipping errors Electronic charge errors Long delay Delivery error
537 2,460 650 5,372 752
Construct a Pareto diagram for these data. What conclusions would you reach? 13. The number of defects found in 25 samples of 100 Gamma Candy Company lemon drops taken on a daily basis from a production line over a five-week period is given here (by rows). Plot these data on a run chart, computing the average value (center line), but ignoring the control limits. Do you suspect that any special causes are present? Why? 0 14 3
544310036 12 17665763 2 2 4 6
Chapter 13
Tools for Process Improvement
679
Figure 13.24 Automobile Service Flowchart (Problem 10)
Stage 1—Preliminary Activities
Stage 3—Perform Work
Stage 2—Problem Diagnosis
Stage 4—Billing and Vehicle Retrieval
Source: Reprinted from "Make Your Service Fail-Safe," by Richard B. Chase and Douglas M. Stewart, Sloan Management Review, 40-41, by permission of the publisher. © 1994 by Sloan Management Review Association. All rights reserved.
14. A pharmaceutical company that manufactures individual syringes is con¬ ducting a process capability study (see Chapter 11). The data shown in the C13dataset file for Prob. 13-14 on the student CD-ROM represent the lengths of 35 consecutive samples.33 Plot these data on a run chart. Do the data appear to come from a stable system so that a process capability study may be conducted appropriately? 15. The Monterey Fiesta Mexican Restaurant is trying to determine whether its popular Pan Con Mucho Sabor breadsticks are correlated with the sales of margaritas. It has data on sales of breadstick baskets and margaritas for 25 weeks, shown in the C13dataset file for Prob. 13-15 on the student CD-ROM. Use the cor¬ relation utility, along with a scatter diagram, in Microsoft Excel to analyze these data. What do they indicate?
680
Part 3
Six Sigma and the Technical System
Figure 13.25 Medical Administration Process (Problem 11)
Source: Ellen Williams and Ray Tailey, "The Use of Failure Mode Effect and Criticality Analysis in a Medication Error Subcommittee," ASQC Health Care Division Newsletter, Winter 1996, 4.
jjjgH
Projects, Etc 1. Research several companies to identify the type of problem-solving approach they use in their improvement efforts. Compare and contrast their approaches. Which, if any, of the approaches described in the chapter are they most similar to? 2. Work with your school administrators to identify an important quality-related problem they face. Outline a plan for improvement. If time permits, apply some of the problem-solving tools to collect data, identify the root cause, and gen¬ erate ideas for solving the problem or improving the situation. 3. Develop a flowchart of the process you use to study for an exam. How might you improve this process? 4. Describe a personal problem you face and how you might use the Deming cycle and the Seven QC Tools to address it. 5. In small teams, develop cause-and-effect diagrams for the following problems: a. Poor exam grade b. No job offers c. Late for work or school 6. Work with teachers at a local high school or grade school to identify some stu¬ dents who are having difficulties in school. Apply quality tools to help find the source of the problems and create an improvement plan.
Chapter 13
Tools for Process Improvement
681
7. Identify several sources of errors as a student or in your personal life. Develop some poka-yokes that might prevent them. 8. Interview a plant manager or quality professional at one or more local compa¬ nies to see whether they use any poka-yoke approaches to mistake-proof their operations. 9. Check out the Web site http://www.freequality.org. It contains descriptions and examples of the use of quality improvement tools. Find some that have not been discussed in this chapter and develop a short tutorial for using them. 10. Examine the ProcessModel output for the help desk example. Are the results compatible with the assumptions given about the process? For example, what percentages of calls actually go to the second support person? Are the simulated times close to the stated input values? 11. Search the Internet for John Grout's Poka-Yoke Web site. Read several of the interesting articles available there and write a report on the information you discover. 12. Meet with key decision makers in your fraternity, sorority, or other organization for which strategic issues need to be considered and prioritized. Research nom¬ inal group technique beyond the scope of this text to have a clear idea of how to proceed. Apply it to your organization's planning and report on the results.
Hj*® Additional cases are available in the Bonus Materials Folder on the CD-ROM. I. Readilunch Restaurant Carole Read, the owner of the Readilunch Restau¬ rant, a downtown, quick service restaurant, was concerned about the loss of several regular cus¬ tomers. She measured the number of empty lunch tables from 11 a.m. until 2 p.m. over a four-week period. To better understand the reasons for the loss of customers, long lines, and dissatisfied patrons, Carol talked to several regular customers. She found that they liked the food and atmosphere of the restaurant, but felt that there were opportu¬ nities for improvement based on the lack of capa¬ bility to quickly handle take-out orders (they had to be phoned in, not faxed), excessive time spent waiting for tables, inefficient service, surly waiters on certain days, and long lines at the cash register. She puzzled over how to sort out possible causes that led to these perceived problems. Carol also decided to design a check sheet to systematically gather data and determine which of these prob¬ lems were the most significant.
Note: Data for the check sheet information gathered for "Vacant Tables" and for "Customer Concerns" can be found in the C13dataset.xls file under the Readilunch 1 and Readilunch 2 tabs on the student CD-ROM. Discussion Questions 1. Plot the average number of empty tables on a run chart, computing the average value (center line), but ignoring the control limits. What do these data show? 2. Use one of the seven Tools to come up with possible causes to explain customer dissatis¬ faction, based on the reasons described in the case. 3. Analyze the check sheet data on the student CD-ROM. What conclusions do you reach? 4. What do you recommend that Carol do to overcome these problems?
682
Part 3
Six Sigma and the Technical System
II. National Furniture
National Furniture is a large retail design and fur¬ niture store. The store often orders special mer¬ chandise at the request of its customers. However, the store recently experienced problems with the on-time delivery of these special orders. Some¬ times the orders were never received, resulting in irate customers. The process of fulfilling a special order begins with the sales associate who records the customer information and obtains approval from a manager to process the order. The sales associate puts the order form in a bin for the office manager to fax to the special order department at the regional office. When the office manager faxes the special order forms from the bin, she files them in a notebook. If a problem arises with the order, the manager receives notification, and contacts the sales asso¬ ciate who took the order to decide what needs to be done next. Typical problems that are often observed include sales associates not filling out the order form completely or entering a request date that is impossible to fulfill. Sometimes the sales associate does not put the form in the proper bin, so the form never gets faxed. Other times, sales associates are asked to obtain more informa¬ tion from the customer, but fail to call the cus-
tomer 'back, or do not inform the office associate to refax the form after getting additional information from the customer. At the regional office, the special order depart¬ ment receives the fax from the store, reviews it, and informs the store if additional information is needed. When all the information is complete, they process the order. Sometimes they lose or misplace the form after it arrives on the fax machine, order the wrong merchandise, or fail to notify the store when additional information is needed or when the merchandise should be expected to arrive. Discussion Questions
1. Develop a process map for special orders. What steps might you suggest to improve this process? 2. Construct a cause-and-effect diagram for identifying reasons why special orders are not received on time. 3. Discuss the relationship between the process map and the cause-and-effect diagram. How can they be used together to attack this problem?
III. Janson Medical Clinic
The Janson Medical Clinic recently conducted a patient satisfaction survey of 100 patients. Using a scale of 1-5, with 1 being "very dissatisfied" and 5 being "very satisfied," the clinic compiled a check sheet for responses that were either 1 or 2, indi¬ cating dissatisfaction with the performance attrib¬ utes. This check sheet is shown in Table 13.5. Doctors have extremely busy schedules. They have surgeries to perform, and many are teaching faculty at the local medical school. Many surgeries are emergencies or take longer than expected, resulting in delays of getting back to the clinic. In the clinic, one or two telephone receptionists answer calls for three different departments, which include 20 or more doctors. Their job is basically to schedule appointments, provide direc¬ tions, and transfer calls to the proper secretaries, which generally requires putting the patient on hold. Often, the receptionist must take a hand¬ written message and personally deliver it to the
secretary because the secretary's phone line is busy. However, the receptionist cannot leave her desk without someone else to cover the phones. A student intern examined the processes for answering phone calls and registering patients. The flowcharts she developed are shown in Fig¬ ures 13.26 and 13.27. Discussion Questions
1. Construct a Pareto diagram for dissatisfac¬ tion. What conclusions do you reach? 2. Select the top three sources of patient dissat¬ isfaction and propose cause-and-effect dia¬ grams for the possible reasons behind them. 3. Propose some process improvements to the flowcharts in Figure 13.26 and develop redesigned processes along with new flow¬ charts. How will your suggestions address the sources of dissatisfaction in Table 13.5?
Chapter 13
Tools for Process Improvement
Table 13.5 Check Sheet of Dissatisfied Responses Making an Appointment Ease of getting through on the phone—10 Friendliness of the telephone receptionist—5 Convenience of office hours—7 Ease of getting a convenient appointment—12
Check-in/Check-out Courtesy and helpfulness of the receptionist—7 Amount of time to register—1 Length of wait to see a physician—13 Comfort of registration waiting area—4
Care and Treatment Respect shown by nurses/assistants—0 Responsiveness to phone calls related to care—5 How well the physician listened—3 Respect shown by the physician—2 Confidence in the physician's ability—1 Explanation of medical condition and treatment—2
Figure 13.26 Current Process for Answering Phone Calls
683
684
Part 3
Six Sigma and the Technical System
Figure 13.27 Current Patient Registration Process
ENDNOTES 1. Gregory Korte, "473 Steps," The Cincinnati Enquirer, October 30, 2002, Al, A10. 2. Gerald Langley, Kevin Nolan, and Thomas Nolan, "The Foundation of Improvement," Sixth Annual Inter¬ national Deming User's Group Conference, Cincinnati, OH (August 1992).
3. Langley et al. (see note 2). 4. Jeremy Main, "Under the Spell of the Quality Gurus," Fortune, August 18,1986,31. 5. Masaaki Imai, Kaizen: The Key to japan's Competi¬ tive Success (New York: McGraw-Hill, 1986), 15. 6. A. F. Osborn, Applied Imagination, 3rd ed. (New
Chapter 13
Tools for Process Improvement
York: Scribner's, 1963); S. J. Parnes, R. B. Noller, and A. M. Biondi (eds.), Guide to Creative Action (New York: Scribner's, 1977). 7. Daniel R. Heiser and Paul Schikora, "Flow¬ charting with Excel," Quality Management Journal 8, no. 3 (2001), 26-35. 8. AT&T Quality Steering Committee, Reengineering Handbook, AT&T Bell Laboratories (1991), 45. 9. Rochelle Rucker, "Six Sigma at Citibank," avail¬ able at http://www.insidequality.wego.net. 10. Adapted from Dwight Kirscht and Jennifer M. Tunnell, "Boise Cascade Stakes a Claim on Quality," Quality Progress 26, no. 11 (November 1993), 91-96. With permission of Dwight M. Kirscht, Timber and Wood Products Division, Boise Cascade Corporation. 11. Kaoru Ishikawa, Guide to Quality Control, 2nd rev. ed. (Tokyo: Asian Productivity Organization, 1986). Available from UNIPUB/Quality Resources, One Water Street, White Plains, NY 10601. 12. Adapted from Bruce Rudin, "Simple Tools Solve Complex Problems." Reprinted with permission from Quality, April 1990, 50-51; a publication of Hitchcock Publishing, a Capital Cities/ABC, Inc. 13. Adapted from Rudin (see note 12). 14. Eleanor Chilson, "Kaizen Blitzes at Magnivision: $809,270 Cost Savings," Quality Management Forum 29, no. 1 (Winter 2003). 15. For an interesting, albeit academic discussion of the psychology of human error and its relationship to mistake-proofing, see Douglas M. Stewart and Richard B. Chase, "The Impact of Human Error on Delivering Service Quality," Production and Operations Management 8, no. 3 (Fall 1999), 240-263; and Douglas M. Stewart and John R. Grout, "The Human Side of Mistake Proofing," Production and Operations Management 10, no. 4 (Winter 2001), 440H59. 16. From Poka-Yoke: Improving Product Quality by Pre¬ venting Defects. Edited by NKS/Factory Magazine, Eng¬ lish translation copyright ©1988 by Productivity Press, Inc., P.O. Box 3007, Cambridge, MA 02140, 800-394-6868. Reprinted by permission. 17. Harry Robinson, "Using Poka-Yoke Techniques for Early Defect Detection," Paper presented at the Sixth International Conference on Software Testing and Analysis and Review (STAR '97). 18. Excerpts reprinted from Richard B. Chase and
685 Douglas M. Stewart, "Make Your Service Fail-Safe," Sloan Management Revieiv 35, no. 3 (Spring 1994), 35-44. © 1994 by the Sloan Management Review Association. All rights reserved. 19. Steve Fleming and E. Lowry Manson, "Six Sigma and Process Simulation," Quality Digest, March 2002. 20. Fleming and Manson (see note 19). 21. This example is adapted from a tutorial for ProcessModel, a commercial simulation package (see note 22). ProcessModel, Inc. 32 West Center, Suite 209, Provo, UT 84601. 22. ProcessModel, Inc. 32 West Center, Suite 209, Provo, UT 84601. 23. James P. Lewis, Team-Based Project Management (New York: Amacom, 1998). 24. Peter R. Scholtes, The Team Handbook, 3rd ed. Madison, WI: Oriel, Inc., 2003, 4-2 through 4-5. 25. Scholtes (see note 24), 4-5. 26. Andre L. Delbecq, Andre H. Van de Ven, and David H. Gustafson, Group Techniques for Program Plan¬ ning (Glenview, IL Scott Foresman and Co., 1975). 27. John E. Bauer, Grace L. Duffy, and Russell T. Westcott (eds.), The Quality Improvement Handbook (Mil¬ waukee, WI: ASQ Quality Press, 2002), 108-109. 28. Adapted from Timothy Clark and Andrew Clark, "Continuous Improvement on the Free-Throw Line," Quality Progress, October 1997, 78-80. © 1997 American Society for Quality. Reprinted with permission. 29. Appreciation is expressed to one of the author's students, Jim Faze, who wrote the paper on which this case is based, as part of the requirements for MGT 640, Total Quality Management, 2001, at Northern Kentucky University. 30. Courtesy of Siemens Energy and Automation Distribution Products Division. 31. Courtesy of Lucas Sumitomo Brakes, Inc., and "Easy Money" team members Ron Gogan, Darren Brown, Jeff Carroll, Mike Watkins, Denis Muse, Marte Wolfensperzjer, and Sean Miller. 32. Adapted from Ronald G. Conant, "JIT in a Mail Order Operation Reduces Processing Time from Four Days to Four Hours," Industrial Engineering 20, no. 9 (September 1988), 34-37. 33. Leroy A. Franklin and Samar N. Mukherjee, "An SPC Case Study on Stabilizing Syringe Lengths," Quality Engineering 12, no. 1 (1999-2000), 65-71.
BIBLIOGRAPHY AT&T Quality Steering Committee. Batting 1000. AT&T Bell Laboratories, 1992. -. Process Quality Management & Improvement Guidelines. AT&T Bell Laboratories, 1987. Box, G. E. R, and S. Bisgaard. "The Scientific Con¬
text of Quality Improvement." Quality Progress 20, no. 6 (June 1987), 54-61. Brassard, Michael. The Memory Jogger Plus+. Methuen, MA: GOAL/QPC, 1989. Evans, James R., and David L. Olson. Introduction to
686 Simulation and Risk Analysis. Upper Saddle River, NJ: Prentice Hall, 2002. Gitlow, H., S. Gitlow, A. Oppenheim, and R. Oppenheim. Tools and Methods for the Improvement of Quality. Homewood, IL: Irwin, 1989. Godfrey, Blan. "Future Trends: Expansion of Quality Management Concepts, Methods and Tools to All Indus¬ tries." Quality Observer 6, no. 9 (September 1997), 40-43, 46. Harrington, H. James, and Kerim Tumay. Simulation Modeling Methods. New York: McGraw-Hill, 2000.
Part 3
Six Sigma and the Technical System
Hradesky, John L. Productivity and Quality Improve¬ ment. New York: McGraw-Hill, 1988. TheTnc Team. The Team Memory Jogger. Madison, WI: Brian Joyner and Associates, Goal/QPC, 1995. Tomas, Sam. "Six Sigma: Motorola's Quest for Zero Defects." APICS, The Performance Advantage (July 1991), 36-41. -. "What Is Motorola's Six Sigma Product Quality?" American Production and Inventory Control Society 1990 Conference Proceedings. Falls Church, VA: APICS, 27-31.
Statistical Process Control QUALITY Profiles: Trident Precision Manufacturing, Inc. and Operations Management Inter¬ national, Inc. Quality Control Measurements Capability and Control
SPC
Methodology
Control Charts for Variables Data
Constructing x- and R-Charts and Establishing Statistical Control Interpreting Patterns in Control Charts Process Monitoring and Control Estimating Process Capability Modified Control Limits Excel Spreadsheet Templates Special Control Charts for Variables Data
Charts for Defects Choosing Between c- and M-Charts Summary of Control Chart Construction Designing Control Charts
Basis for Sampling Sample Size Sampling Frequency Location of Control Limits SPC, ISO 9000:2000, and Six Pre-Control
Quality in Practice: Applying spc to Pharmaceutical Product Manufacturing Quality in Practice: Using a u-Chart in a Receiving Process
x- and s-Charts
Review Questions
Charts for Individuals
Problems
Control Charts for Attributes
Fraction Nonconforming (p) Chart Variable Sample Size
Sigma
Controlling Six Sigma Processes
CASES: La Ventana Window Company Murphy Trucking, Inc. Day Industries
np-Charts for Number Nonconforming
Deming's funnel experiment, described in Chapter 11, demonstrates that failure to distinguish between common causes and special causes of variation can actually increase the variation in a process. This problem often results from the mistaken belief that whenever process output is off target, some adjustment must be made. Knowing when to leave a process alone is an important step in maintaining control over a process. Equally important is knowing when to take action to prevent the pro¬ duction of nonconforming product.
687
Part 3
688
Six Sigma and the Technical System
Statistical process control (SPC) is a methodology for monitoring a process to identify special causes of variation and signaling the need to take corrective action when it is appropriate. When special causes are present, the process is deemed to be out of control. If the variation in the process is due to common causes alone, the process is said to be in statistical control. A practical definition of statistical control is that both the process averages and variances are constant over time.1 SPC relies on control charts, one of the basic quality improvement tools that we briefly introduced in Chapter 13. SPC is a proven technique for improving quality and productivity. Many customers require their suppliers to provide evidence of sta¬ tistical process control. Thus, SPC provides a means by which a firm may demon¬ strate its quality capability, an activity necessary for survival in today's highly competitive markets. Because SPC requires processes to show measurable variation, it is ineffective for quality levels approaching six sigma. However, SPC is quite effec¬ tive for companies in the early stages of quality efforts. Although control charts were first developed and used in a manufacturing con¬ text, they are easily applied to service organizations. Table 14.1 lists just a few of the many potential applications of control charts for services. The key is in defining the appropriate quality measures to monitor. Most service processes can be improved through the appropriate application of control charts. In this chapter we describe how to develop and use statistical process control to monitor manufacturing and service processes. The Bonus Materials folder for this chapter provides an explanation of the statistical details for understanding the theory underlying control charts.
Table 14.1 Control Chart Applications in Service Organizations Quality Measure
Organization Hospital
Lab test accuracy Insurance claim accuracy On-time delivery of meals and medication
Bank
Check-processing accuracy
Insurance company
Claims-processing response time Billing accuracy
Post Office
Sorting accuracy Time of delivery Percentage of express mail delivered on time
Ambulance
Response time
Police Department
Incidence of crime in a precinct Number of traffic citations
Hotel
Proportion of rooms satisfactorily cleaned Checkout time Number of complaints received
Transportation
Proportion of freight cars correctly routed Dollar amount of damage per claim
Auto service
Percentage of time work completed as promised Number of parts out of stock
Chapter 14
Statistical Process Control
689
Quality Profiles Trident Precision Manufacturing, Inc. and Operations Management Trident
International, Inc. Founded in 1979 with only three people, pri¬ Headquartered in Greenwood Village, Colorado, vately held Trident manufactures precision sheet Operations Management International, Inc. metal components, electromechanical assem¬ (OMI), operates and maintains more than 160 blies, and custom products, mostly in the office- public and private sector wastewater and water equipment, medical supply, computer, and treatment facilities in 29 states and facilities in defense industries with a workforce of about Brazil, Canada, Egypt, Israel, Malaysia, New 170. Trident's human resource strategies empha¬ Zealand, Philippines, and Thailand. OMI's pri¬ size training, involvement through teams, mary services are processing raw wastewater to empowerment, and reward and recognition. produce clean, environmentally safe effluent and Since 1989, Trident has invested 4.4 percent of its processing raw groundwater and surface water to payroll in training and education, two to three produce clean, safe drinking water. OMI's "E3" times the average for all U.S. industry and an motto, "Exceed our customers' expectations, especially large amount for a small firm. empower our employees, enhance the environ¬ All goals, however, contribute to achieving ment," is the foundation for its Quality as a Busi¬ Trident's overarching aim of total customer satis¬ ness Strategy leadership system. Key enablers are faction. Each improvement project begins with a the company's Linkage of Process Model, which thorough analysis of how to meet or exceed cus¬ defines relationships among processes, and its tomer requirements in four critical areas: quality, Family of Measures, a balanced scorecard of 20 cost, delivery, and service. Metrics are designed integrated metrics. These measures of operational to ensure that progress toward the customer- performance correspond to OMI's four strategic targeted improvements can be evaluated. The objectives—customer focus, business growth, company's data collection system provides all innovation, and market leadership. personnel with a current record of the company's Improvement initiatives in the company's progress toward its goals. The senior executive strategic plan are selected and crafted so that each team also reviews performance data in each initiative contributes significantly to achieving department daily and weekly. Once each month, one or more strategic objectives and key customer this team aggregates the data for the entire com¬ requirements. In 2000, OMI had 26 improvement pany and reports on progress toward goals set initiatives under way, each one assigned to a team for each of the five key business drivers. led by a high-level executive. All teams write char¬ Beyond tracking its operational and financial ters that state their purpose, objectives, and time¬ performance, Trident also analyzes data collected line for completion. A team charter also specifies from a variety of other internal and external which of OMI's more than 150 critical processes sources. These sources include semiannual sur¬ are involved, the metrics that will be used for eval¬ veys of customers, suppliers, and employees; uation, costs, required resources, and other infor¬ benchmarking studies; discussions with custo¬ mation vital to the success of the initiative. mers; employee forums; market reports; quarterly Charters provide team members and company quality audits; and an independently conducted executives with the means for a quick and thor¬ annual assessment of the company's competitive ough analysis of progress toward planned goals. position within its industry. Defect rates have OMI received a 2000 Baldrige Award. fallen so much that Trident offers a full guarantee Source: Malcolm Baldrige National Quality Award, Profiles of against defects in its custom products. Trident was Winners, National Institute of Standards and Technology, Depart¬ a 1996 recipient of the Baldrige Award. ment of Commerce.
Part 3
690
Six Sigma and the Technical System
QUALITY CONTROL MEASUREMENTS Quality control measurements and indicators fall into one of two categories. An attribute is a performance characteristic that is either present or absent in the product or service under consideration. For example, a dimension is either within tolerance or out of tolerance, an order is complete or incomplete, or an invoice can have one, two, three, or any number of errors. Thus, attributes data are discrete and tell whether the characteristic conforms to specifications. Attributes can be measured by visual inspection, such as assessing whether the correct zip code was used in shipping an order; or by comparing a dimension to specifications, such as whether the diameter of a shaft falls within specification limits of 1.60 ± 0.01 inch. Attribute measurements are typically expressed as proportions or rates, for example, the fraction of noncon¬ formances in a group of items, number of defects per unit, or rate of errors per oppor¬ tunity. The second type of performance characteristic is called a variable. Variables data are continuous (e.g., length or weight). Variables measurements are concerned with the degree of conformance to specifications. Thus, rather than determining whether the diameter of a shaft simply meets a specification of 1.60 ± 0.01 inch, a measure of the actual value of the diameter is recorded. Variable measurements are generally expressed with such statistics as averages and standard deviations. Table 14.1 provides addi¬ tional examples of both attributes and variables measurements. In a statistical sense, attributes inspection is less efficient than variables inspection; that is, it
Collecting attribute data is usually easier than collecting variable data because the assessment can usually be done more quickly by a simple inspection or count, whereas vari¬ able data require the use of some type of measuring instrument.
does not provide as much information. This means that attributes inspection requires a larger sample than variables inspection to obtain the same amount of statistical information about the quality of the product. This difference can become significant when inspection of each item is time-con¬ suming or expensive. Most quality characteristics in services are attributes, which is perhaps one reason why service organizations have been slow to adopt measure¬ ment-based quality management approaches.
CAPABILITY AND CONTROL Consider Table 14.2, which shows measurements of a quality characteristic for 30 samples from a manufacturing process with specifications 0.75 ± 0.25. Each row corre¬ sponds to a sample size of 5 taken every 15 minutes. The mean of each sample is also given in the last column. A frequency distribution and histogram of these data is shown in Figure 14.1. The data form a relatively symmetric distribution with a mean of 0.762 and standard deviation 0.0738. Using these values, we find that Cpk - 1.075, indicating that the process capability is at least marginally acceptable. Because the data were taken over an extended period of time, we cannot deter¬ mine whether the process remained stable. In a histogram, the dimension of time is not considered. Thus, histograms do not allow you to distinguish between common and special causes of variation. It is unclear whether any special causes of variation are influencing the capability index. If we plot the mean of each sample against the time at which the sample was taken (because the time increments between samples are equal, the sample number is an appropriate surrogate for time), we obtain the run chart shown in Figure 14.2. It indicates that the mean has shifted up at about sample 17. In fact, the process average for the first 16 samples is only 0.738 while the average
Chapter 14
Statistical Process Control
691
Table 14.2 Quality Measurements for Thirty Samples A 1 Sample 2 1 J 2 4 3 5 4 6 5 7 6 8 7 9 8 10 9 11 10 12 11 13 12 14 13 15 14 16 15 17 16 18 17 19 18 20 19 21 20 22 21 23 22 24 23 25 24 26 25 27 26 28 27 29 28 30 29 31 30
B 0.682 0.787 0.780 0.591 0.693 0.749 0.791 0.744 0.769 0.718 0.787 0.622 0.657 0.806 0.660 0.816 0.826 0.828 0.805 0.802 0.876 0.855 0.762 0.703 0.737 0.748 0.826 0.728 0.803 0.774
C
D Observations 0.689 0.776 0.860 0.601 0.667 0.838 0.727 0.812 0.708 0.790 0.714 0.738 0.713 0.689 0.779 0.660 0.77.3 0.641 0.671 0.708 0.821 0.764 0.802 0.818 0.822 0.893 0.749 0.859 0.681 0.644 0.817 0.763 0.77? 0.721 0.829 0.865 0.719 0.612 0.756 0.786 0.803 0.701 0.783 0.722 0.705 0.804 0.337 0.759 0.723 0.776 0.686 0.856 0.803 0.764 0.721 0.820 0.892 0.740 0.837 0.872
E
F
0.798 0.746 0.785 0.775 0.758 0.719 0.877 0.737 0.644 0.850 0.658 0.872 0.544 0.801 0.747 0.716 0.770 0.778 0.938 0.815 0.789 0.856 0.805 0.975 0.748 0.811 0.823 0.772 0.816 0.849
0.714 0.779 0.723 0.730 0.671 0.606 0.603 0.822 0.725 0.712 0.708 0.727 0.750 0.701 0.728 0.649 0.809 0.872 0.807 0.801 0.672 0.751 0.80.9 0.732 0.732 0.838 0.886 0.639 0.770 0.818
G
H Mean 0.732 0.755 0.759 0.727 0.724 0.705 0.735 0.748 0.710 0.732 0.748 0.768 0.733 0.783 0.692 0.753 0.781 0.834 0.776 0.792 0.768 0.793 0.777 0.801 0.743 0.788 0.820 0.736 0.804 0.830
for the remaining samples is 0.789. Therefore, although the overall average is close to the target specification, at no time was the actual process average centered near the target. We should conclude that this process is not in sta¬ tistical control, and we should not pay much attention to the process capability calculations. Control and capability are two different concepts. As shown in Figure 14.3, a process may be capable or not capable, or in con¬ trol or out of control, independently of each other. Clearly, we would like every process to be both capable and in control. If a process is neither capable nor in con¬ trol, we must first get it in a state of control by removing special causes of variation, and then attack the common causes to improve its capability. If a process is capable but not in control (as the previous example illustrated), we should work to get it back in control. Process capability calculations make little sense if the process is not in sta¬ tistical control because the data are confounded by special causes that do not represent the inherent capability of the process.
Part 3
692
Six Sigma and the Technical System
Figure 14.1 Frequency Distribution and Histogram B
A 1 2 3 4
D
E
F
G
Frequency distribution Upper limit Frequency 0.55
1
0.6
5
0.65
1 10
6 7
0.7 0.75
40
8
0.8
31
9
0.85
37
10
0.9
14
11 12
0.95
1 1
13 I More
1
Histogram
14
14 15 Average 16 Std. Dev.
0.762 0.0738
17
Bin
18 19
Figure 14.2 Run Chart of Sample Means
Sample
Chapter 14
Statistical Process Control
693
Figure 14.3 Capability versus Control (Arrows indicate the direction of appropriate management action) Control In Control
Out of Control
Capability
Capable
Not Capable
SPC METHODOLOGY Control charts, like the other basic tools for quality improvement, are relatively simple to use. Control charts have three basic applications: (1) to establish a state of statistical control, (2) to monitor a process and signal when the process goes out of control, and (3) to determine process capability. The following is a summary of the steps required to develop and use control charts. Steps 1 through 4 focus on estab¬ lishing a state of statistical control; in step 5, the charts are used for ongoing moni¬ toring; and finally, in step 6, the data are used for process capability analysis. 1. Preparation a. Choose the variable or attribute to be measured. b. Determine the basis, size, and frequency of sampling. c. Set up the control chart. 2. Data collection a. Record the data. b. Calculate relevant statistics: averages, ranges, proportions, and so on. c. Plot the statistics on the chart. 3. Determination of trial control limits a. Draw the center line (process average) on the chart. b. Compute the upper and lower control limits. 4. Analysis and interpretation a. Investigate the chart for lack of control. b. Eliminate out-of-control points. c. Recompute control limits if necessary. 5. Use as a problem-solving tool a. Continue data collection and plotting. b. Identify out-of-control situations and take corrective action. 6. Determination of process capability using the control chart data
Part 3
694
Six Sigma and the Technical System
In the remainder of this chapter we discuss the construction, interpretation, and use of control charts following this methodology. Although many different charts are described, they differ only in the type of measurement for which the chart is used; the methodology previously described applies to each of them.
CONTROL CHARTS FOR VARIABLES DATA Variables data are those that are measured along a continuous scale. Examples of vari¬ ables data are length, weight, time, and distance. The charts most commonly used for variables data are the x-chart ("x-bar" chart) and the R-chart (range chart). The x-chart is used to monitor the centering of the process, and the R-chart is used to monitor the variation in the process. The range is used as a measure of variation simply for conve¬ nience, particularly when workers on the factory floor perform control chart calcula¬ tions by hand. For large samples and when data are analyzed by computer programs, the standard deviation is a better measure of variability (discussed later in this chapter).
Constructing x- and /?-Charts and Establishing Statistical Control The first step in developing x- and R-charts is to gather data. Usually, about 25 to 30 samples are collected. Samples between size 3 and 10 are generally used, with 5 being the most common. The number of samples is indicated by k, and n denotes the sample size. For each sample i, the mean (denoted x,) and the range (R,j are computed. These values are then plotted on their respective control charts. Next, the overall mean and average range calculations are made. These values specify the center lines for the x- and Rcharts, respectively. The overall mean (denoted x) is the average of the sample means x,-: k
k The average range (R) is similarly computed, using the formula: it
k The average range and average mean are used to compute upper and lower control limits (UCL and LCL) for the R- and x-charts. Control limits are easily calculated using the following formulas:
UCLR = D4R
UCLf = x + a2r
LCLr = D,R
LCL? = x - A2R
where the constants D3, D4, and A2 depend on the sample size and can be found in Appendix B. The control limits represent the range between which all points are expected to fall if the process is in statistical control. If any points fall outside the control limits or if any unusual patterns are observed, then some special cause has probably affected
Chapter 14
Statistical Process Control
695
the process. The process should be studied to determine the cause. If special causes are present, then they are not representative of the true state of statistical control, and the calculations of the center line and control limits will be biased. The corresponding data points should be eliminated, and new values for x, R, and the control limits should be computed. In determining whether a process is in statistical control, the R-chart is always analyzed first. Because the control limits in the x-chart depend on the average range, special causes in the R-chart may produce unusual patterns in the x-chart, even when the centering of the process is in control. (An example of such distorted patterns is given later in this chapter.) Once statistical control is established for the R-chart, attention may turn to the x-chart. Figure 14.4 shows a typical data sheet used for recording data and drawing con¬ trol charts, which is available from the American Society for Quality (ASQ). This form provides space for descriptive information about the process, recording of sample observations and computed statistics, and drawing the control charts. On the back of this form is a worksheet (see Figure 14.5) for computing control limits and process capability information. The construction and analysis of control charts is best seen in an example. The thickness of silicon wafers used in the production of semiconductors must be carefully controlled. The tolerance of one such product is specified as ±0.0050 inches. In one production facility, three wafers were selected each hour and the thick¬ ness measured carefully to within one ten-thousandth of an inch. Figure 14.6 on page 698 shows the results obtained for 25 samples. For example, the mean of the first sample is
41 + 70 + 22
113
3
3
44
The range of sample 1 is 70 - 22 = 48. {Note: Calculations are rounded to the nearest integer for simplicity.) The calculations of the average range, overall mean, and control limits are shown in Figure 14.7 on page 699. The average range is the sum of the sample ranges (676) divided by the number of samples (25); the overall mean is the sum of the sample averages (1,221) divided by the number of samples (25). Because the sample size is 3, the factors used in computing the control limits are A2 = 1.023 and D4 = 2.574. (For sample sizes of 6 or less, factor D3 = 0; therefore, the lower control limit on the range chart is zero.) The center lines and control limits are drawn on the chart in Figure 14.8 on page 700. Examining the range chart first, it appears that the process is in control. All points lie within the control limits and no unusual patterns exist. In the x-chart, however, sample 17 lies above the upper control limit. On investigation, some defective mate¬ rial had been used. This data point should be eliminated from the control chart cal¬ culations. Figure 14.9 on page 701 shows the calculations after sample 17 was removed. The revised center lines and control limits are shown in Figure 14.10 on page 702. Customarily, out-of-control points are noted on the chart. The resulting chart appears to be in control. When a process is in statistical con¬ trol, the points on a control chart fluc¬ tuate randomly between the control limits with no recognizable pattern.
Interpreting Patterns in Control Charts The following list provides a set of general rules for examining a process to determine whether it is in control:
696
Part 3
Figure 14.4 ASQ Control Chart Data Sheet
DC
ofl IX DC
< X
o J o DC I— 2 O o cf) HI
_i
ffi
< DC
< >
C\l
CO
'T
in
lx
siN3i/\iaansv3iAj 31dlAJVS
Source: Reprinted with permission of ASQ.
Six Sigma and the Technical System
Chapter 14
Statistical Process Control
697
Figure 14.5 ASQ Control Chart Calculation Worksheet CALCULATION WORKSHEET CONTROL LIMITS
LIMITS FOR INDIVIDUALS
SUBGROUPS INCLUDED _
COMPARE WITH SPECIFICATION OR TOLERANCE LIMITS
R- — R k *
EX
d2 "
"
x “ ~iT ULy = X + lR X 1 (MIDSPEC. OR STD.) LL A2R =
= X -
x
2 R d2
US UCLj = X + A2R
LS
LCL5 - X-A2R UCLp = D„R -
US - LS x
6a =
2r d2
MODIFIED CONTROL LIMITS FOR AVERAGES
FACTORS FOR CONTROL LIMITS
BASED ON SPECIFICATION LIMITS AND PROCESS CAPABILITY. APPLICABLE ONLY IF: US-LS > 6c. US
LS
_ A„R =
x
URLy = US - A..R
A„R LRL* = LS + AmR
-
n
A2
D„
d2
d2
2 3 4 5 6
1.880 1.023 0.729 0.577 0.483
3.268 2.574 2.282 2.114 2.004
1.128 1.693 2.059 2.326 2.534
2.659 1.772 1.457 1.290 1.184
0.779 0.749 0.728 0.713 0.701
Source: Reprinted with permission of ASQ.
1. 2. 3. 4.
No points are outside control limits. The number of points above and below the center line is about the same. The points seem to fall randomly above and below the center line. Most points, but not all, are near the center line, and only a few are close to the control limits.
The underlying assumption behind these rules is that the distribution of sample means is normal. This assumption follows from the central limit theorem of statistics, which states that the distribution of sample means approaches a normal distribution as the sample size increases regardless of the original distribution. Of course, for small sample sizes, the distribution of the original data must be reasonably normal for this assumption to hold. The upper and lower control limits are computed to be three stan¬ dard deviations from the overall mean. Thus, the probability that any sample mean falls outside the control limits is small. This probability is the origin of rule 1. Because the normal distribution is symmetric, about the same number of points fall above as below the center line. Also, since the mean of the normal distribution is the median, about half the points fall on either side of the center line. Finally, about 68 percent of a normal distribution falls within one standard deviation of the mean; thus, most—but not all—points should be close to the center line. These characteris¬ tics will hold provided that the mean and variance of the original data have not changed during the time the data were collected; that is, the process is stable. Several types of unusual patterns arise in control charts, which are reviewed here along with an indication of the typical causes of such patterns.2
698
Part 3
Figure 14.6 Silicon Wafer Thickness Data
Six Sigma and the Technical System
Chapter 14
Statistical Process Control
699
Figure 14.7 Control Limit Calculations
CALCULATION WORKSHEET CONTROL LIMITS
LIMITS FOR INDIVIDUALS
SUBGROUPS INCLUDED
COMPARE WITH SPECIFICATION OR TOLERANCE LIMITS
R = '
=
sx_
x
k
a.. HLL
- 27
25 /ZZI IS
4-R □2
= V?8
UL = X +
OR
X‘ (MIDSPEC. OR STD.)
= 50
a2r = I. 023
=
X
21
3d2. R
LL = f - 1 R
*
27. fc
d2
US UCL; - % + AZR
= II. M
LCLj = K - A2R
=
UCL„ = d4r -2.57*/ X
27
LS
2/-2
US - LS =
6
