Six Sigma for Electronics Design and Manufacturing

six sigma, sản xuất

Six Sigma for Electronics Design and Manufacturing

Six Sigma for Electronics Design and Manufacturing

Sammy G ShinaUniversity of Massachusetts, Lowell

McGraw-HillNew York Chicago San Francisco Lisbon London Madrid

Mexico City Milan New Delhi San Juan Seoul

Singapore Sydney Toronto

Copyright © 2002 by The McGraw Hill Companies,Inc All rights reserved Manufactured in the United States of America Except as permitted under the United States Copyright Act of 1976, no part of this publication may be repro- duced or distributed in any form or by any means, or stored in a database or retrieval system, without the prior writ- ten permission of the publisher

The material in this eBook also appears in the print version of this title:0-07-139511-3

All trademarks are trademarks of their respective owners Rather than put a trademark symbol after every rence of a trademarked name, we use names in an editorial fashion only, and to the benefit of the trademark owner, with no intention of infringement of the trademark Where such designations appear in this book, they have been printed with initial caps

occur-McGraw-Hill eBooks are available at special quantity discounts to use as premiums and sales promotions, or for use in corporate training programs For more information, please contact George Hoare, Special Sales, at george_hoare@mcgraw-hill.com or (212) 904-4069

engi-THE WORK IS PROVIDED “AS IS” McGRAW-HILL AND ITS LICENSORS MAKE NO GUARANTEES OR WARRANTIES AS TO THE ACCURACY, ADEQUACY OR COMPLETENESS OF OR RESULTS TO BE OBTAINED FROM USING THE WORK, INCLUDING ANY INFORMATION THAT CAN BE ACCESSED THROUGH THE WORK VIA HYPERLINK OR OTHERWISE, AND EXPRESSLY DISCLAIM ANY WAR- RANTY, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO IMPLIED WARRANTIES OF MER- CHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE McGraw-Hill and its licensors do not warrant

or guarantee that the functions contained in the work will meet your requirements or that its operation will be terrupted or error free Neither McGraw-Hill nor its licensors shall be liable to you or anyone else for any inaccu- racy, error or omission, regardless of cause, in the work or for any damages resulting therefrom McGraw-Hill has

unin-no responsibility for the content of any information accessed through the work Under unin-no circumstances shall McGraw-Hill and/or its licensors be liable for any indirect, incidental, special, punitive, consequential or similar damages that result from the use of or inability to use the work, even if any of them has been advised of the possi- bility of such damages This limitation of liability shall apply to any claim or cause whatsoever whether such claim

or cause arises in contract, tort or otherwise.

DOI: 10.1036/0071409556

abc

McGraw-Hill

0-07-140955-6

Want to learn more?

We hope you enjoy this McGraw-Hill eBook!

If you’d like more information about this book, its author, or related books and websites,

please click here

and our children and grandchildren

to Other Quality Tools

They Effect Six Sigma

1.11.2 Failure modes and effects analysis (FMEA) 261.12 Gauge Repeatability and Reproducibility (GR&R) 29

Their Determination

2.1 The Quality Measurement Techniques: SQC,

2.1.1 The Statistical quality control (SQC) methods 342.1.2 The relationship of control charts and 35six sigma

vii

Copyright 2002 The McGraw-Hill Companies, Inc Click Here for Terms of Use.

For more information about this book, click here.

2.1.5 Six sigma and the 1.5
shift 41

2.3.3 Attribute processes and reject analysis for 57six sigma

Always Normally Distributed?

2.4.2 Checking for normality using chi-square tests 602.4.3 Example of 2goodness of fit to normal 62distribution test

2.4.4 Transformation data into normal distributions 63

normality analysis

3.1 Manufacturing Variability Measurement and Control 70

Relationship with Six Sigma

3.2.4 X
, R variable control chart calculations 76example

3.2.5 Alternate methods for calculating control 78limits

3.2.6 Control chart guidelines, out-of-control 78conditions, and corrective action

procedures and examples

calculations and their relationship tosix sigma

Six Sigma

3.3.1 The binomial distribution 853.3.2 Examples of using the binomial distribution 86

3.3.4 Examples of using the Poisson distribution 873.3.5 Attribute control charts limit calculations 88

calculations and their relationship to six sigma

3.3.7 Use of control charts in factories that are 91approaching six sigma

Quality in Manufacturing

Manufacturing Yield and Test Strategy

Operation or a Part with Multiple Similar Operations4.2.1 Example of calculating yield in a part with 105multiple operations

4.2.2 Determining assembly yield and PCB and 106product test levels in electronic products

4.3.2 Example of yield calculations at the PCB 110assembly level

4.3.3 DPMO methods for standardizing defect 112measurements

4.3.6 The use of implied Cpk in product and 116assembly line manufacturing and

planning activities4.3.7 Example and discussion of implied Cpk in 118

IC assembly line defect projections

4.4.1 PCB test strategy 121

4.4.4 Factors affecting test operation parameters 128

Low-Volume Products and Processes

Calculations for Samples and Populations5.1.1 Examples of the use of the t-distribution for 137sample and population averages

5.1.2 Other statistical tools: Point and interval 138estimation

5.1.3 Examples of point estimation of the average 1395.1.4 Confidence interval estimation for the average 140

populations

determination

production5.2.2 Determination of standard deviation
for 148process capability

5.2.4 Process capability for low-volume production 1505.2.5 Moving range (MR) methodologies for low 150volume: MR control charts

5.2.6 Process capability studies in industry 152

Capability

5.4.1 Process capability for prototype and early 165production parts

5.4.2 Corrective action for process capability 168problems

Electronics Products

6.1 The Overall Electronic Product Life Cycle Cost Model 1706.1.1 The use of the quality and cost model to 173achieve world-class cost and quality

6.1.2 Developing the background information cost 174estimating of electronic products

6.1.3 Determination of costs and tracking tools 176for electronic products

6.2.3 A practical quality and cost approach 181

6.3.1 Relating quality data to manufacturing six 184sigma or Cpk levels

6.3.2 Printed circuit board (PCB) fabrication 185technologies

fabrication cost, and quality issues6.3.4 PCB fabrication cost and quality alternative 191example

6.4.1 Material-based PCB assembly cost system 193

7.2.1 Steps in conducting a successful DoE 211experiment

orthogonal arrays

7.2.6 Multilevel arrangements and combination 225designs

example: Bonding process optimization

7.3.3 Analysis of DoE data with interactions: 232Electrical hipot test L8 partial factorial

Resolution IV example

7.3.5 Statistical analysis of the hipot experiment 236

Manufacturing Projects

and Manufacturing for Current and New

Products and Processes

8.1.1 Process improvement in current products 246

Six Sigma

manufacturing processes8.2.3 Investigating more capable processes for 255new products

investigations for manufacturing: Stencil technology for DoE

8.3 Determining Six Sigma Quality in Different Design 260Disciplines

8.3.2 Mechanical design and tolerance analysis 261

8.3.4 Statistical tolerance analysis for mechanical 263design

8.3.6 Statistical analysis of the mechanical design 265example

processes

example8.3.11 Six sigma for electronic circuits with 270multiple specifications

8.3.12 Special considerations in Cpk for design of 271electronic products

8.3.13 The use of design quality analysis in systems 272architecture

Introduction8.4.1 Optimizing new manufacturing processes 273

Target value manipulations and variabilityreduction DoE

8.4.3 Trade-offs in new product design disciplines 2778.4.4 New product design trade-off example— 277Screening DoE followed by in-depth DoE for defect elimination in thermal printer design

9.1 Background: Concurrent Engineering Successes 288and New Trends

9.1.1 Changes to the product realization process 291

9.2.4 Supply chain communications and 300information control

Quality Issues

9.3.2 Changing traditional design communications 310and supplier involvement

Using Six Sigma Quality

10.1 The Quality System Review and Quality-Based 318Project Management Methodologies

10.1.1 The quality-based system design process 31810.1.2 Six sigma quality-based system design 319process benefits

management

development process10.2 Technical Design Information Flow and Six Sigma 327System Design

10.2.1 Opportunities in six sigma for system or 328product design improvements

10.2.7 Standardized procedures in manufacturing 335

to determine the composite Cpk

11.1.1 Axioms for creating six sigma within 340the organization

11.2 Cultural Issues with the Six Sigma Based System 348Design Process

Product Creation Process

quality systems review11.3.3 Six sigma manufacturability assessment 353and tactical plans in production

Illustrations and Tables

Illustrations

Figure 1.1 World-class benchmarks percentage improvements per year 5

Figure 2.3 Intersection of process capability and specification 37

limits to determine the defect level

Figure 2.4 Conceptual view of control and capability concepts 38

Figure 2.9 Graphical presentation of normal distribution with parts 53

compliance percentage and multiple
limits

Figure 2.12 Quick visual check for normality in Example 2.4.1 61

Figure 2.14 Plot of observed (dark) versus expected (clear) frequencies 64Figure 2.15 Plot of Example 2.4 data set original (top) and transformed 65

by –log x
on the bottom.

Figure 4.2 An example of a multistep manufacturing process line 108

xvii

Copyright 2002 The McGraw-Hill Companies, Inc Click Here for Terms of Use.

Figure 4.3 DPMO chart example 115

Figure 5.3 Confidence interval around the mean  and
is known 141

Figure 5.5 Obtaining confidence limits from 2distribution with 144

confidence (1 – )%

Figure 5.9 Distributions of prototype and early production of parts 166

Figure 6.2 Typical cost distribution of an electronic product 176Figure 6.3 Cost history of an electronic product based on the 177

concept stage

Figure 6.4 Volume sensitivity of the cost of an electronic product 178

Figure 6.7 A typical approach to printed circuit board (PCB) assembly 194

Figure 7.3 The use of an L8 as full factorial versus saturated design 219Figure 7.4 The plot of interactions of the example in Table 7.6 222Figure 7.5 Linear graphs for the interactions of L8 shown in Table 7.7 223

Figure 8.1 Progression of quality tools for existing products 245Figure 8.2 Cause and effect diagram for mixed technology PCBs 248Figure 8.3 Graphical analysis of DoE for mixed technology soldering of 248

PCBs

Figure 8.4 Histogram of solder defects distribution 6 months before 250

and after DoE

Figure 8.5 Overall new product quality, including design and 251

manufacturing

Figure 8.8 Mechanical design of a typical vibrating angioplasty probe 267Figure 8.9 S/N analysis for fine pitch SMT processing variability 276

Figure 9.3 Traditional versus concurrent engineering project 294

communications

Figure 9.5 Supplier management models 302

Figure 10.2 Development project time line: phases and milestones 326

Tables

Table 1.1 Criteria rating (CR) to select a solder system for PCB assembly 12

Table 2.1 Defect rates in PPM for different quality levels and 41

distribution shifts

Table 4.2 Yield calculation in a line with n parts in a three-step 110

production line

Table 4.3 DPMO grouping of defects and opportunities for PCB 112

assemblies

three strategies

Table 5.1 Selected values of t,of student’s t distribution 137

Table 5.4 Amount of data required for process capability studies 146Table 5.5 Example of process capability studies for PCB assembly line 154

fabrication shop

Table 6.4 Classifications for different types of PCB assemblies 193

Table 6.6 Cost rate calculations for machine-loaded TH components 197Table 6.7 Technology cost model with modifiers for PCB assembly 198Table 6.8 Cost model drivers example for sheet metal fabrication 200

Table 7.2 XOR logic table for interaction level determinations 216

Table 7.7 Interaction scenarios for L8 with Resolution IV design 223

Table 7.14 F table value for 95% confidence or 0.05 confidence 236

Table 7.16 Hipot design ANOVA statistical analysis with pooled error 237Table 8.1 Design and analysis of DoE for mixed technology PCBs 247

Table 8.3 Simulation results for Cpk analysis of a bandpass filter 253Table 8.4 Quality data for PCB assembly manufacturing processes 254Table 8.5 Quality analysis of a two-sided PCB with TH, SMT, and 255

mechanical assembly and multiple components and leadsTable 8.6 Quality drivers for printed circuit board (PCB) assembly 256Table 8.7 DoE stencil technology experiment factor and level selection 258

Table 8.9 Stencil technology percent contribution analysis of average 259

solder deposition area

Table 8.10 Stencil technology quality loss function (QLF) formula 260Table 8.11 Tolerance analysis for three-part example, worst-case 264

analysis

Table 8.12 Tolerance analysis for three-part example, six sigma analysis 265

Case 3: statistical tolerance

Table 8.23 Printer in-depth DoE analysis and final recommnedations 282Table 9.1 Status of companies outsourcing hardware design and 290

manufacturing capabilities

Table 9.2 Common issues in selecting outsourced products and 296

competencies

Table 9.5 Weighted quality criteria for supplier selection matrix 306

Table 9.7 Attributes and metrics of success for each design phase 311Table 9.8 Changes from traditional engineering to new methodologies 314

Table 10-1 Total product development process concept-to-development 322

criteria

specifications and modes

Abbreviations

AQAP Advance product quality planning and control plan

ANOVA Analysis of variance

BIST Built in self-test

BOM Bill of materials

CEM Contract electronic manufacturers

Cpk Capability of the process, with average shift

DFT Design for testability

DPMO Defect per million opportunities

ERP Enterprise requirements planning

ESI Early supplier involvement

IPC Institute for Interconnecting and Packaging of Electronic Circuits FMEA Failure mode effect analysis

GR&R Gauge repeatability and reproducibility

MTBF Mean time between failure

xxiii

Copyright 2002 The McGraw-Hill Companies, Inc Click Here for Terms of Use.

NS Normal (probability) score

RFI Radio frequency interference

TQC Total quality control

Preface

Six sigma is becoming more important as companies compete in aworldwide market for high-quality, low-cost products Successful im-plementations of six sigma in different companies, large and small, donot follow identical scripts The tools and methodologies of six sigmaare fused with the company’s culture to create a unique and success-ful blend in each instance

This book is intended to introduce and familiarize design, tion, quality, and process engineers and their managers with many ofthe issues regarding the use of six sigma quality in design and manu-facturing of electronic products, and how to resolve them It is based

produc-on my experience in practicing, cproduc-onsulting, and teaching six sigmaand its techniques over the last 15 years During that time, I confront-

ed many engineers’ natural reservation about six sigma: its tions are too arbitrary, it is too difficult to achieve, it works only forlarge companies, it is too expensive to implement, it works only formanufacturing, not for design, and so on They continuously chal-lenged me to apply it in their own areas of interest, presenting mewith many difficult design and manufacturing six sigma applicationproblems to solve At the same time, I was involved with many compa-nies and organizations whose engineers and managers were usingoriginal and ingenious applications of six sigma in traditional designand manufacturing Out of these experiences came many of the exam-ples and case studies in this book

assump-I observed and helped train many engineers in companies usingtools and methodologies of six sigma The companies vary in size,scope, product type, and strategy, yet they are similar in their ap-proach to successfully implementing six sigma through an interdisci-plinary team environment and using the tools and methods men-tioned in this book effectively by altering them to meet theirparticular needs

I believe the most important impact of six sigma is its use in the sign of new products, starting with making it one of the goals of thenew product creation process It makes the design engineers extreme-

de-ly cognizant of the importance of designing and specifying productsthat can be manufactured with six sigma quality at low cost Toomany times, a company introduces six sigma by having manufactur-

xxv

Copyright 2002 The McGraw-Hill Companies, Inc Click Here for Terms of Use.

ing adopt it as its goal, a very daunting task, especially if currentproducts were not designed with six sigma in mind.

The approach I use in this book is not to be rigid about six sigma Ihave attempted to present many of the options available to measureand implement six sigma, and not to specifically recommend a course

of action in each instance Engineers are very creative people, andthey will always try to meld new concepts into ones familiar to them.Many will put their own stamp on its methodology or add their ownway of doing things to the six sigma techniques The one sure way tomake them resist a new concept is to force it down their throats I be-lieve these individual engineers’ efforts should be encouraged, as long

as they do not detract from the overall goal of achieving six sigma

I hope that this book will be of value to the neophyte as well as theexperienced practitioners of Six Sigma In particular, it will benefitthe small to medium size companies that do not have the support staffand the resources necessary to try out some of the six sigma ideas andtechniques and meld them into the company culture The experiencesdocumented here should be helpful to encourage many companies toventure out and develop new world-class products through six sigmathat can help them grow and prosper for the future

I am indebted to several organizations for supporting and ing me during the lengthy time needed to collect my materials, writethe chapters, and edit the book I thank The University of Massachu-setts, Lowell for its continuing support for product design and manu-facturing, especially Chancellor Bill Hogan; the Dean of the James B.Francis College of Engineering, Krishna Vedula; and the chairman ofthe Department of Mechanical Engineering, John McKelliget TheReed Exhibition Companies and SMTA, through their NEPCON andSMTI conferences in Anaheim and Chicago, encourage and nurturethe design and manufacturing of electronic products

encourag-In addition, I offer my thanks to Mr Steve Chapman of Hill, Inc., who was my editor for this book, as well as my previous twobooks on concurrent engineering He always believed in me and en-couraged and guided me through three books, and for that I am verygrateful

McGraw-Finally, many thanks to my family for emotional support during thewriting, editing, and production of the book, including my wife Jackieand our children, Mike, Gail, Nancy, and Jon, as well as my grand-children, who brought me great joy between the many days of writingand editing I also thank the many attendees of my seminars on sixsigma and quality methods, including the in-company presentations,who kept alive my interest and faith in six sigma I wish them success

in implementing six sigma tools and methods in their companies.Sammy G Shina

January, 2002

1

The Nature of Six Sigma

and Its Connectivity

to Other Quality Tools

The modern attention to the use of statistical tools for the ture of products and processes originated prior to and during WorldWar II, when the United States of America geared up to a massivebuildup of machinery and arms to successfully conclude the war Theneed to manage the myriad of complex weapon systems and their var-ied and distributed defense contractors led to the evolution of the sys-tem of Statistical Quality Control (SQC), a set of tools that culminat-

manufac-ed in the military standards for subcontracting, such as MIL-Std 105.The term “government inspector” became synonymous with those in-dividuals who were trained to use the tables that controlled theamount of sampling inspection between the different suppliers ofparts used by the main weapons manufacturers The basis of the SQCprocess was the use of 3 sigma limits, which yields a rate of 2700 de-fective parts per million (PPM)

Prior to that period, large U.S companies established a qualitystrategy of vertical integration In order to maintain and managequality, companies had to control all of the resources used in the prod-uct Thus, the Ford Motor Company in the early part of the 20th cen-tury purchased coal and iron mines for making steel for car bodiesand forests in Brazil to ensure a quality supply of tires This strategywas shelved during the rapid buildup for the war because of the use ofcoproducers as well as subcontractors

1

Copyright 2002 The McGraw-Hill Companies, Inc Click Here for Terms of Use.

The war was won and U.S companies returned to their originalstrategy while the defeated countries were rebuilding their industries.

In order to revive the Japanese economy, General McArthur, who wasthe governor general of Japan at that time, imported some of the U.S.pioneers of SQC to help train their counterparts in Japan These effortswere largely successful in transforming Japanese industry from a low-technology producer of low-quality, low-cost products such as toys tothe other side of the spectrum By the 1970s and 1980s Japanese prod-ucts were renowned for their quality and durability Consumers andcompanies flocked to buy Japanese electronics, cars, and computerchips, willing to pay a premium for their high quality In recognition ofthis effort, Japan established the Deming prize for quality, which waslater emulated in the United States, with the Baldrige award

U.S companies’ response to their loss of market share to Japanesecompanies was to investigate the Japanese companies’ secrets of suc-cess Many U.S companies organized trips in the 1980s to Japanesecompanies or branches of U.S companies in Japan Initial findingswere mostly unsuccessful Japanese concepts such as “quality circles”

or “zero defects” did not translate well into the U.S companies’ ture Quality circles, which were mostly ad hoc committees of engi-neers, workers, and their managers, were created to investigate qual-ity problems In many cases, they were not well organized, and aftermany months of meetings and discussions, resulted in frivolous solu-tions It was also difficult to implement quality circles in unionizedshops The term zero defects was also ambiguous, because it was hard

cul-to define: Does the fact that a production line produces a million partsand only one is found to be defective constitute a failure to reach thezero defects goal?

The industrial and business press in the 1980s was filled with cles comparing Japanese and U.S quality The pressures mounted toclose the quality gap U.S Companies slowly realized that quality im-provements depended on the realization of two major elements—theyhave to be quantifiable and measurable, and all elements that makethe company successful must be implemented: superior pricing, deliv-ery, performance, reliability, and customer satisfaction All of thecompany’s elements, not just manufacturing, have to participate inthis effort, including management, marketing, design, and external(subcontractors) as well as internal suppliers (in-house manufactur-ing) The six sigma concept satisfies these two key requirements,which has led to its wide use in U.S industry today

arti-The Motorola Company pioneered the use of six sigma Bill Smith,Motorola Vice President and Senior Quality Assurance Manager, is

widely regarded as the father of six sigma He wrote in the Journal of Machine Design issue of February 12, 1993:

For a company aiming to design products with the lowest possible ber of defects, traditional three-sigma designs are completely inade-quate Accordingly in 1987, Motorola engineers were required to createall new designs with plus or minus six sigma tolerance limits, giventhat the sigma is that of a world-class part or process in the first place.This marked the start of Motorola’s Six Sigma process and its adoption

num-of robust design as one capable num-of withstanding twice the normal tion of a process

varia-Early in 1987, Bob Galvin, the CEO of Motorola and head of itsOperating/Policy Committee, committed the corporation to a planthat would determine quality goals of 10 times improvement by

1989, 100 times improvement by 1991, and six sigma capability by

1992 At that time, no one in the company knew how to achieve thesix sigma goal, but, in their drive for quality, they committed thecompany to reach the six sigma defect rate of just 3.4 defective partsper million (PPM) in each step of their processes By 1992, they metthese goals for the most part At several Motorola facilities, theyeven exceeded six sigma capability in some products and processes

On average, however, their manufacturing operations by 1992 were

at about 5.4 sigma capability, or 40 defective PPM—somewhat short

of their original goal

The six sigma effort at Motorola has led to a reduction of in-processdefects in manufacturing by 150 times from 1987 to 1992 Thisamounts to total savings of $2.2 billion since the beginning of the sixsigma program Richard Buetow, Motorola’s Director of Quality, com-mented that six sigma reduced defects by 99.7% and had saved thecompany $11 billion for the nine-year period from 1987 to 1996.Today, Motorola has reached its goal of six sigma The complexity ofnew technology has resulted in a continued pressure to maintain thishigh level of quality As product complexity continues to increase—such as semiconductor chips with billions of devices and trillions of in-structions per second—it will be essential that Motorola master theprocess of producing quality at a parts-per-billion level That is quite

a challenge One part per billion is equivalent to one second in 31years!

Therefore, Motorola expanded the six sigma program in 1992 andbeyond to achieve the following:

1 Continue their efforts to achieve six sigma results, and beyond, ineverything they do

2 Change metrics from parts per million to parts per billion (PPB)

3 Go forward with a goal of 10 times reduction in defects every 2years

Many other companies have also adopted these high levels of

quali-ty, as well as cost reduction, responsiveness, flexibiliquali-ty, and inventoryturnover One of the most notable is the General Electric Company(GE) Several GE executives commented on the six sigma program in

an article by Rachel Lane, a reporter for Bloomberg news, in 1997 and

in the GE annual report for the same year James McNerney, CEO of

GE Aircraft Engines said:

Foremost among our initiatives, Six Sigma Quality is driving culturalchange throughout our entire operation and accelerating our businessresults Six Sigma tools allow us to improve results dramatically by en-hancing the value we provide to our customers Almost one third of ouremployees have been trained to lead projects and spread Six Sigmatools to co-workers, resulting in more than $70 million in productivitygains in 1997

The same year, GE Appliance Director/CEO David Cote said: “This is

a leap of faith, when people see the actual results that come from thisand make money, you think, ‘Son of a gun, this thing really doeswork!’ ”

Jeffery Immelt, CEO of GE Medical Systems said in 1997: “If youwant to change the way you do things, you have to have people whoare in the game.” To that end, GE created a class of six sigma practi-tioners that take their titles from the martial arts Extensive Train-ing was provided to all employees Those at the top were called “blackbelts” and “master black belts.” They work on six sigma full time andassist in training and leading six sigma projects Regular employeeswho receive abridged training are called “green belts.”

During the last few decades, advances in the high-technology and tronics industries have accelerated The price/performance ratios con-tinue to follow the industry idioms of more performance for lower price.Intel’s Gordon Moore first proposed the law that bears his name in thelate 1960s: chip complexity (as defined by the number of active ele-ments on a single semiconductor chip) will double about every devicegeneration, usually about 18 calendar months This law has now beenvalid for more than three decades, and it appears likely to be valid forseveral more device generations The capacity of today’s hard drives isdoubling every nine months; and the average price per megabit havedeclined from $11.54 in 1988 to an estimated $0.02 in 1999

elec-Great expansion has also been occurring in the field of tion, both in the speed and the availability of the Internet It is esti-mated that that global access to the Internet has increased from 171

communica-million people in March 1999 to 304 communica-million in March 2000, an crease of 78%

in-In quality, similar improvements have been made, as shown by some

of the numbers quoted above These improvements have led to an crease in customer expectations of quality Companies have responded

in-to this increase by continuously measuring themselves and their petition in several areas of capabilities and performance This concept,also known as benchmarking, is a favorite tool of managers to set goalsfor the enterprise that are commensurate with their competition Theycan also gauge the progress of enterprises toward achieving their goals

com-in quality, as well as cost, responsiveness, flexibility, and com-inventoryturnover Figure 1.1 is a spider diagram of U.S versus world classbenchmarks outlining annual improvements generated by Motorola in

1988, showing the range of capabilities and their annual percentageimprovements over a 4 year average period At that time, it was esti-mated that the average business in the United States is somewhatprofitable, with market prices declining and new competitors enteringthe marketplace These companies were spending 10–25% of sales dol-lars on reworking defects Concurrently, 5–10% of their customerswere dissatisfied and would not recommend that others purchase theirproducts These companies believed that typical six sigma quality isneither realistic nor achievable, and were unaware that the “best inclass” companies are 100 times better in quality

The inner closed segment in Figure 1.1 represents an average U.S.company in 1988, profiled above The middle segment represents aworld class company, and the outer segment represents the best inclass companies World class is the level of improvements that is

1000 800 600 400 200

Best in ClassWorld Class

Figure 1.1 World-class benchmarks, percentage improvements per year.

needed to compete globally Best in class represents the best able annual improvements recorded anywhere, and not necessarily inthe business segment that the company competes in It is the bench-mark of what is achievable in any measure of performance.

achiev-It is apparent that an accurate method for developing and ing quality systems in design and manufacturing as well as customersatisfaction is needed to achieve these high quality and capability re-sults, and to compete with products that can be designed, manufac-tured, and sold anywhere Six sigma is an excellent tool to achieveworld class status as well as best in class results in quality, especiallygiven the increased complexity of designs and products

improv-At the same time, the requirements for developing new products inhigh-technology industries have followed these increases in complexi-

ty and improvements in quality, necessitating faster product ment processes and shorter product lifecycles Many of the leadingtechnology companies have created “virtual enterprises,” aligningthemselves with design and manufacturing outsourcing partners tocarry out services that can be performed more efficiently outside theboundaries of the organization These partnerships enabled a compa-

develop-ny to focus on its core competencies, its own product brand, its tomers, and its particular competency in design or manufacturing These newly formed outsourcing companies are providing cost-effective and timely services In manufacturing, they provide multi-disciplinary production; test and support services, including printedcircuit board (PCB) assembly and testing and packaging technologysuch as sheet metal and plastic injection molding; and software con-figuration and support services such as repair depot and warranty ex-changes They also offer lower cost, higher flexibility, and excellentquality, eliminating the need to spend money on capital equipmentfor internal capacity This new outsourcing model allows all links inthe supply chain to focus on their own core competencies while stillreducing overall cycle times

cus-In design outsourcing, the supply chain offers the flexibility of gle or multiple competencies, including specialized engineering analy-sis and design validation, testing, and conformance to design stan-dards for multiple countries or codes In addition, suppliers can offertheir own supply chain of strategic alliances in tooling and manufac-turing services worldwide Most of these outsourcing companies offerdesign feedback in terms of design for manufacture (DFM) throughearly supplier involvement (ESI) These design service providers havereduced the need for high-technology companies to purchase or main-tain expensive engineering and design competencies, such as specificdesign analysis, some of which are used infrequently in project designcycles

sin-Several industries, especially the auto industry, have worked tostandardize their relationship with their suppliers They created theAdvance Product Quality Planning (APQP) Task Force Its purposewas to standardize the manuals, procedures, reporting format, andtechnical nomenclature used by Daimler-Chrysler, Ford, and GeneralMotors in their respective supplier quality systems for their designand manufacturing The APQP also issued a reference manual devel-oped by the Measurement Systems Analysis (MSA) Group for insur-ing supplier compliance with their standards, especially QS9000.These standards contain many of the principles of six sigma and asso-ciated quality tools, such as Cpk requirements These manuals werepublished in the mid-1990s and are available from the Automotive In-dustry Action Group (AIAG) in Southfield Michigan

Six sigma can be used as a standard for design and manufacturing,

as well as a communication method between design and ing groups, especially when part of the design or manufacturing isoutsourced This is important for companies in meeting shorter prod-uct lifecycles and speeding up product development through faster ac-cess to design and manufacturing information and the use of globalsupply chains

Six sigma, like many new trends or initiatives, is not without its ics and detractors The author has run into several issues brought up

crit-by engineers and managers struggling with six sigma concepts, andhas attempted to address these concerns by writing this book Some ofthe most frequent critiques of six sigma, and the author’s approach toaddressing these problems are listed below

1 The goal of six sigma defects, at 3.4 PPM, and some of its ples, such as the ±1.5 sigma shift of the average manufactured partfrom specification nominal, sound arbitrary In addition, there is nosolid evidence as to why these numbers have been chosen

princi-These are reasonable assumptions that were made to implement sixsigma There are other comparable systems, such as Cpk targets used

in the auto industry, that could substitute for some of these tions Discussions of these concepts are in Chapters 2 and 3

assump-2 The cost of achieving six sigma might result in a negative return

on investment Conventional wisdom once held that higher qualitycosts more, or that there is an optimum point at which cost and quali-

ty balance each other, and any further investment in quality will sult in negative returns (see the discussion of the quality loss function

re-in Chapter 6)

These beliefs are based on the misconceptions that more tests and

inspections are needed in the factory prior to delivery to the customer,

in order to deliver higher quality Six sigma advocates the tion of these costs during the design stage, prior to the manufacturingrelease of the product, so that these costs are well understood In ad-dition, it has been demonstrated in six sigma programs that the cost

identifica-of changing the product in the design stage to achieve higher quality,whether through design changes, different specifications, better man-ufacturing methods, or alternate suppliers, are much lower than sub-sequent testing and inspection in manufacturing These issues arediscussed in the chapters on product testing (Chapter 4) and cost(Chapter 6)

3 Many companies feel that the six sigma programs only work wellfor large-volume, well-established, and consumer-oriented companiessuch as Motorola and GE, but do not work for other industries such asaerospace, defense, or medical, since their volumes are small or theyare more focused on maximizing the performance of products or re-ducing the time of development projects

There are many statistical methods that can be used to supplantthe sampling and analysis required for six sigma, allowing smallercompanies the full benefits of six sigma in product design and manu-facturing Six sigma methods can be used successfully to introducenew low-volume products as well as quantifying marginal designs.These methods will be discussed in the chapters on high and low vol-ume (Chapter 5) and six sigma current and new products (Chapter 8)

4 Many engineers feel that six sigma is for manufacturing only,not for product design, and that it is very difficult to accomplish andcannot be achieved in a timely manner

In this book, there will be many examples of using six sigma and itsassociated tools, such as design of experiments (DoE), in product de-sign These methods can help in realizing the six sigma goals and tar-gets in a timely and organized manner in design and manufacturing

In addition, there are many examples where design engineers weresurprised to find out that they are already achieving six sigma in cur-rent designs Six sigma can also be used to flush out “gold plated” de-signs: designs that are overly robust, beyond the six sigma limits, andtherefore costing more than required These issues are discussed inChapter 7 on DoE and Chapter 8 on designing current and new prod-ucts

Six sigma integrates well with all of the quality programs and trends

of the last few decades The purpose of this section is to outline ceptually where the six sigma program connects in the quality hierar-

con-chy and some of the quality tools that are in common use today cific mathematical background and formulations are discussed in de-tail in later chapters.

Spe-Six sigma is a condition of the generalized formula for process bility, which is defined as the ability of a process to turn out a goodproduct It is a relationship of product specifications to manufacturingvariability, measured in terms of Cp or Cpk, or expressed as a numer-ical index Six sigma is equivalent to Cp = 2 or Cpk = 1.5 (more onthat in the next chapter) The classical definition of the capability ofthe process or Cp is:

Six sigma is achieved when the product specifications are at ±6
(

is the symbol for standard deviation) of the manufacturing processcorresponding to Cp = 2 (or Cpk = 1.5, discussed in Chapter 2)Six sigma or Cp is an excellent indicator of the capability of aprocess, which can be expressed numerically This numerical expres-sion can be translated into a defect level using normal distributionstatistical assumptions It is a useful tool for manufacturing processcomparisons, as well as a common language of design and manufac-turing personnel during the development phase of a product The de-sign project team and their managers can use it to set new productquality goals It can be used to assess the quality of internal manufac-turing plants anywhere in the world or to measure the capability of asupplier Companies can use it to communicate a particular contrac-tual level of quality for their supply chain

The quality tools in wide use today can easily be integrated within thesix sigma definitions The object of six sigma is to steadily increasethe process capability index until it reaches the desired level: thespecification limits of the design are equal to six sigma of the manu-facturing variability

product specifications



manufacturing variability

USL – LSL

6
(total process range from –3
to +3
)

specification width (or design tolerance)



process capability (or total process variation)

Design engineers normally set the product specifications, whereasmanufacturing engineers are responsible for production variability.The object of increasing the process capability to six sigma or Cp = 2 istwofold: increase the product specifications, either by widening them

or reducing the manufacturing variability Either effort can have apositive effect on reaching six sigma

The design specifications for any part or process are related to thetop published product specifications Ultimately, it is the customerthat determines the relative importance of each specification and thedesired level of performance Good market research and project man-agement for new products can determine the best level of specifica-tion This level can be set to balance the wishes of the customer, tem-pered by what the competition is offering and considering inputs fromdesign and manufacturing engineers as to the difficulty of meetingthat specification level

The quality of supplied parts and the efforts of the manufacturingengineers in production solely determine the denominator of the sixsigma equation, or the manufacturing variability Implementing thetraditional quality tools of manufacturing, such as statistical qualitycontrol (SQC) and associated quality tools, can reduce the manufac-turing variability The tools of SQC and their relationship to processcapability are discussed in Chapter 3

The Cp formula can then be rewritten as

They Affect Six Sigma

Before a six sigma effort is launched, it is mandatory to have a defined and successfully managed total quality management (TQM)program The tools of TQM encourage the use of well-establishedmethodologies for quantifying, analyzing, and resolving quality prob-lems A brief description of the TQM tools and examples of each will

well-be given in Chapter 2

It is widely recognized that TQM tools and techniques should be infull utilization before the launch of any six sigma program SQCshould also be well implemented in the organization, with wide use ofcontrol charts in manufacturing and the supply chain Both of thesetools will be discussed in Chapter 3 regarding process control

customer

supplier

The major tools of quality can be arrayed as to their use in ing six sigma Some tools can effect the numerator, denominator, orboth elements in Equation 1.4 However, a definition of each majortool is given below, in order to examine its relationship with six sig-ma.

QFD is a structured process that provides a means for identifying andcarrying the customer’s voice through each stage of product develop-ment and implementation QFD is achieved by cross-functional teamsthat collect, interpret, document, and prioritize customer require-ments to identify bottlenecks and breakthrough opportunities QFD is a market-driven design and development process resulting

in products and services that meet or exceed customer needs and pectations It is achieved by hearing the voice of the customer, direct-

ex-ly stated in their own words, as well as anaex-lyzing the competitive sition of the company’s products and services Usually, a QFD team isformed, consisting of marketing, design, and manufacturing engi-neers, to help in designing new products, using customer inputs andcurrent product capabilities as well as competitive analysis of themarketplace QFD can be used alternately for new product design aswell as focusing the efforts of the QFD team on improving existingproducts and processes QFD combines tools from many traditionaldisciplines, including engineering, management, and marketing

po-1.8.1 Engineering

Tools such as structured analysis or process mapping, which is a down division of requirements into multiple elements in severalcharts, each related to a requirement in the higher chart, are em-ployed An example of two tiers of structured analysis is given in Fig-ures 1.6 and 1.7 and will be discussed later in this chapter

top-1.8.2 Management

Tools such as decision analysis (DA) or criteria rating (CR) are ployed This technique consists of breaking a complex decision intodistinct criteria, ranking each alternative decision versus each criteri-

em-on, then adding the total weighted criteria to determine the most fective overall decision An example of criteria rating is the decision

ef-on a soldering material for PCB assembly given in Table 1.1 Thereare four alternatives being considered by the selection team, and thecriteria for the decision are listed on the left side of Table 1.1, each