JVIANAGERIAL

Techniques for Improving Quality and

Productivity in the Workplace

DEVELOP
A RELIABLE
METHOD

DRIVING
POWER
FOR
MANAGERIAL
KEEP
SUCCESS
EVERY CREATE A
WORKER FAVORABLE
PRACTICED IN ENVIRONMENT
^HE METHOD

RYUII FUKUDA
Foreword by Norman Bodek
MANAGERIAL
ENGINEERING
MANAGERIAL
ENGINEERING
Techniques for Improving Quality and
Productivity in the Workplace

RYUJI FUKUDA
Foreword by Norman Bodek
President. Productivity, Inc.

PRODUCTIVITY, INC.
Stamford, Connecticut
Copyright © 1983 by Productivity, Inc.

Originally published by the Japanese Standards Association in 1982 as

— QC. IE. OR ©«»9<);sffl;4 —

Library of Congress Catalog Card No. 83-62993

ISBN 0-915299-00-3

All rights reserved. No part of this publication may be reproduced or transmitted

in any form or by any means without permission in writing from the publisher.

Productivity, Inc.
PO Box 16722
Stamford, CT 06905

Excerpts from Loud and Clear Tf\e Full Answer to Aviation's Vital Question; Are ]ets Really

Safe? byRobert H. Serling. Copyright ® 1969 by Robert H. Serling Reprinted by

permission of Doubleday & Company. Inc.

Printed in the United States of America

First English edition 6 5 4 3 2

Set in Novarese
Printed and bound by Halliday Lithograph.
CONTENTS

List of Figures ix

List of Tables xiii

Foreword xv

Acknowledgements xix

Introduction xxi

CHAPTER I

Principles of Managerial Engineering 1

1.1 The Search for a Reliable Method 1

1.2 Creating a Favorable Environment 4

1.3 Practicing Together 10

CHAPTER 2

2.1 Lessons Learned fronn Accidents 13

2.2 Driven by Necessity 15
2.3 Extracting the Essence 16
2.4 Taking a New Look at Human Errors 17
2.5 The Day the "Gray Workshop" Changed 20
Appendix to Chapter 2 22

CHAPTER 3

A Reliable Method ( 1
)-. The Cause and Effect

Diagram with the Addition of Cards 27

3.1 Ideal Conditions 28
3.2 The Quality Control Problem Study Group —
Conditions for Development 29
Vi MANAGERIAL ENGINEERING

3.3 What is CEDAC:' 33

3.4 Steps for Implementing the CEDAC System 36
3.5 Some Variations on CEDAC 42
3 The Conceptual Framework of CEDAC
6 43
3.7 Unexpected Results 44
3.8 Lessons from the Hawthorne Study 46
3.9 The Hypothesis of New foharry's Window 47
3.10 Defense (Adhering to Established Standards)
and Offense (Finding New Methods) 51
3.11 Creating a CEDAC Model 57
3.12 Grouping of the Questionnaire Items 59
3.13 Participation by All Leads to a Breakthrough 64
3.14 Conclusions from CEDAC Applications 67
3.15 QC for the Constant Production of High Quality
Products 69
3.16 Thorough Practice 70
3.17 Quality Cost 71
3.18 A Professor from Denmark 75
3.19 Necessary Conditions for Reliable Methods 78

CHAPTER 4
A ReliableMethod 2 Stockless Production
( ) : 8
4. 1 Production Control —
Basics for the
Manufacturer 81
4.2 Developing Theories Out of Practice 82
4.3 Moving Towards Stockless Production 84
4.4 Production Standards as Management Policy 87
4.5 The Characteristic Approach 91
4.6 Ebb Tide at the Seashore 91
4.7 Building a Basic Model 94
4.8 Approaches to Stockless Production 102
4 9 How Stockless Production Was Developed 106
4.10 Second Thoughts on OR 107
 Contents vii

CHAPTER 5

A Reliable Method (3): IE Improvements 1 1

5.1 Robots and IE 111

Reducing Setup Time
'

5.2 113
5.3 The "Bunt" Strategy 113
5.4 Improving Main Operations 120
5.5 Achieving Improvement Through Practice 120
5.6 Points on the Observation of Operations 1 23
5.7 Principles for the Improvement of Operations 1 25
5.8 An Improvement
Invitation to IE 129
5.9 The Contribution of Worker Group Activities 130
Appendix to Chapter 5 1 37

CHAPTER 6
Two Approaches to "Practice By All" 141
6.1 OET: Utilizing Errors for Practice 141
6.2 Practicing OC by the Case Method 148

CHAPTER 7
Situational Analysis. Find Your Problems Yourself 1 53
7.1 Cartoons for Equipment Maintenance 153
7.2 Replacing Formal Meetings 154
7.3 Clarifying the Problem Situation 155
7.4 Detect Your Own Problems 158
7.5 Correct Situational Analysis and Correct Actions 159

CHAPTER 8
An Invitation to Self-Study — For the Active Mind 163
8. On the Hawthorne Study 163
8.2 Improving the Company's Constitution 169

Postscript 173

Index 175
FIGURES

Figure 1-1 Driving Power for Management

Activities facing page 1

Figure 1-2 Results of Work Improvement 6

Figure 1-3 Productivity Growth at Company E 9
Figure 3-1 Cause-and-Effect Diagram 34
Figure 3-2 Basic Structure of CEDAC 36
Figure 3-3 Process of Incorporating Data on Cards 38
Figure 3-4 Some Variations of CEDAC 40
Figure 3-5 The Basic Conceptual Framework
of CEDAC 43
Figure 3-6 loharry'sWindow 48
Figure 3-7 New loharry's Window 49
Figure 3-8 Finding New Methods and Adhering
to Established Standards 53
Figure 3-9 Groups Which Found Effective
Methods as a Direct Result of CEDAC 53
Figure 3-10 Drawing Up CEDAC: Adhering to
Standards Gives Rise to New Methods 55
Figure 3-11 Drawing Up CEDAC: Necessary
Conditions for Adhering to Standards 56
Figure 3-12 Formulation of the CEDAC Model 58
Figure 3-13 Decrease in Quality Defect Rate 59
Figure 3-14 Scores of the Attributes of Production
Environment 61
 MANAGERIAL ENGINEERING

Figure 3-15 Scores of the Attributes of

Psychological Environment 62
Figure 3-16 The Adherence-improvement Cycle 68
Figure 3-17 Workers List Problems Causing Defects
in Rolled Steel Wire Operation 76
Figure 3-18 Scientific Methodology 79
Figure 4-1 Work-in-Process Reduction 86
Figure 4-2 Manufacturing Lead Time Reduction 86
Figure 4-3 Productivity Improvement 86
Figure 4-4 Standard Flow Time 88
Figure 4-5 Work-in-Process and Production
Problems 92
Figure 4-6 Simulation Form 96
Figure 4-7 Manufacturing Man-Hours and
Lead Time 100
Figure 4-8 Results of Limiting Work-In-Process
Material 101

Figure 4-9 Methods for Achieving Stockless

Production 104
Figure 5-1 An Example of Operation Analysis 112
Figure 5-2 Cases as Classified by the Amount of
Time Saved 117
Figure 5-3 Body Motions and Average Time
Required 122
Figure 5-4 Accumulation of Small Improvements
by Large Workforce 125

Figure 5-5 A Regression Model for Efficiency

Growth 133
Figure 5-6 Contributions to MH Efficiency Growth 134

Figure 5-7 Basic Framework for Productivity

Improvement 135
 Figures xi

Figure 6-1 Trend of Work-In-Process Quality Defects 142

Figure 6-2 Analysis of Failure by OET
(lanuary-February 1978) 144
Figure 6-3 Analysis of Failure by OET
(lanuary-February 1979) 144
Figure 6-4 Four Essential Factors for OET 146
Figure 7- 1 Ideas on TPM Developed by Workers
Using Cards 156
Figure 8-1 Production System Improvement
Through IE Activities 164
Figure 8-2 Meiden's Window 168
Figure 8-3 Form for Categories 2 and 3 170
Figure 8-4 Form for Category 4 170
Figure 8-5 Summary Table 171
TABLES

Table 2-1 Potential Dangers in an Operation 19

Table 3-1 Purposes of Standardization 29

Table 3-2 Results of Applying CEDAC 45
Table 3-3 Patterns of CEDAC Activities: Where the
Defect Rates Decreased 66
Table 3-4 How QC Activities Lead to Quality
Improvenaent 69
Table 3-5 Realizing a Zero-Defect Production System:
A Summary of the QC Problems Study
Group's ideas 72
Table 3-6 Quality Cost Data for Plant A 74
Table 4-1 A Comparison of Two Control Systems 90
Table 4-2 Reasons for Work-in-Process Accumulation 93
Table 4-3 Principles for Production Control 95
Table 4-4 Lanchester's Law 104
Table 5-1 Basic Steps for Setup Time Reduction 114
Table 5-2 Shortening of internal Setups 116
Table 5-3 Shortening a Die Change Operation 118
Table 5-4 Principles of Motion Economy 127
Table 5-5 Problems and the IE Course Curriculum 132
Table 6-1 Comparison of QET and Previous Group
Activities 147
FOREWORD

I first met Ryuji March 1981 while was leading my

Fukuda in I

first industrial study mission to lapan. During this two week

tour we met with managers from sixteen of Japan's leading cor-

porations. Our group of nineteen American senior executives
was really fired up about discovering the secrets of lapanese
management. What truly lay behind the "miracle" of lapan's
quality and productivity gains? What was kanban? How effec-
tive, really, were quality control circles, and how did they fit in

with the management culture? Above all, were the techniques being
used here transportable to American soil, or were they specific to the
society of Japan?
We spent about half a day at each company we visited.
During this time we had very little opportunity to get to know
our hosts. They tried to be helpful, but for most of the trip, it
was difficult to overcome the barriers of language and culture
in such a short time.
The one exception was Ryuji Fukuda, at that time a senior
executive with Sumitomo Electric Industries Ltd.

XV
XVi MANAGERIAL ENGINEERING

He did much more than go through the formalities; he ob-

viously wanted very much to teach us. He wanted us to under-
stand that even though we spoke a different language, even
though we looked different to each other, human nature was
the same all overthe world, and that if wecould really see this,
then we would be able to bring the lessons of lapanese man-
agement back with us to the United States.
Ryuji had tremendous barrier to overcome. He didn't
know us personally. He couldn't speak English, and at that
time he had very little translated material we could read about
his work.
However, was aware that he had won the Deming litera-
I

ture prize. felt like a diamond prospector who discovers a for-

1

tune, but only sees before his eyes the cloudy stones. Some-
thing inside me told me that the diamond was there and that
somehow was required to do the polishing.
I

found out that Ryuji could read English, and started to

1

correspond with him. Then invited him to speak at my next

1

"Productivity the American Way" conference, which was to be

held in New York City November, 98 1 1

By November, Ryuji was working with Meidensha Electric

Manufacturing Co. Ltd.. a subsidiary of Sumitomo. The presi-
dent of Meidensha had asked the president of Sumitomo for
help on quality. He specifically asked for Ryuji, who was ap-
pointed to the board of directors and put in charge of quality in
all their production facilities.
Ryuji came to New York, where his presentation was very
well received, and began to understand the full power behind
1

his work. As he said to me then; "1 am

not teaching quality con-
trol; I am teaching foolproof methods for reducing and
"

eliminating defects
was at this time that made a commitment to translate
It I

his work into English and help him teach his concepts to
American managers.
 Foreword xvii

CEDAC is the process that earned Ryuji the Deming prize.

Like many of the other nnethods he teaches in this book, there
is a magic in CEDAC. It is the magic of utter simplicity. Most of

what you will read you already know, but have never been able
to put into practice. As did on my initial journey to japan, you
I

must go deeply into the teaching. You will find the words easy
to read and understand, but you must work very hard with the
author. You can do that by making a point of looking for ways
to apply the principles to your own work.
In this book you will really learn how to apply your own
knowledge. Ryuji is teaching you how to mobilize all of your
experience as a manager and develop for yourself a clear path
to managerial success.
There is no mystery to Japanese management. Most of the
basic principles were discovered in the United States many
years ago. Ryuji has only looked at them and put them to work
scientifically.
I have read this book many times, and each reading leads
me to new discoveries. wish the same for all of Ryuji Fukuda's
I

American readers.
Norman Bodek
Publisher
Productivity, Inc.
ACKNOWLEDGEMENTS

The English verson of Managerial Engineering is realized by

the courtesy of the Japanese Standards Association, an organi-
zation promoting industrial standardization, which is
for
under the control of the Agency of industrial Science and
Technology, Ministry of International Trade and Industry of
Japan (MITl).
The author is grateful for the assistance and support pro-
vided by Sumitomo Electric Industries Ltd. and Meidensha
Electric Manufacturing Co. Ltd.
The publisher wishes to acknowledge the assistance of
the following people: Noriko Hosoyamada, translator; David
Perlstein, editor; Russ Funkhouser, book cover designer; Marie
Kascus, indexer; and Patricia Slote, production manager.

XIX
INTRODUCTION

Foreigners share a widespread misunderstanding that every-

thing runs quite smoothly inJapanese plants. This misun-
derstanding must be corrected, in fact, much time and effort
were needed to achieve the present success of Japan's leading
companies. For example, worker group activities, which have
become famous worldwide, did not work at all fifteen years
many as
ago. Even now, as one-third of the major lapanese
manufacturers cannot benefit from employee participation.
It is important for us to supply unbiased information on

lapanese plant management. I fear that foreigners may other-

wise take lapanese methods as a mysterious formula for qual-
ity and They may mistakenly conclude that lapan
productivity.
is where any problem is solved magically.
a mysterious land
Such misunderstanding obscures the actual efforts of
Japanese engineers and workers which should be of interest to
foreign firms. The mistaken notion that Japan is a peculiar,
mysterious and completely different nation from others does
not provide the basis for true friendship and cooperation, but
rather, danger of international conflict, it is not difficult to

XXI
xxii MANAGERIAL ENGINEERING

avoid such misunderstanding; telling the simple truth will

suffice.
To apply the principles of Japanese plant management in
overseas plants with different social and cultural environ-
ments, we have to analyze our approach to management in
terms of two categories, namely, that which is inherited from
traditional Japanese culture and values, and that which is uni-
versally applicable. If the former factors predominate, the ap-
plication of Japanese plant management in a foreign country
will be impossible.on the other hand, the latter factors
If,

dominate, Japanese management methods can be helpful, if

not wholly applicable.
1 am
convinced that the basic methods of quality control,
industrial engineering and operations research are applicable
in all plants, Japanese and non-Iapanese alike.

In November 98
1 was invited to a three-day conference
1 , 1

in New York sponsored by Productivity, Inc. With the theme

"Productivity - The American Way," the conference brought to-

gether about forty speakers and 400 participants, for the most
part American. They came from various business organiza-
tions, schools, government, and even the Army and Navy.
As the only foreign speaker, presented my ideas on man-
I

agerial engineering with actual case studies from our study

group approach to CEDAC, OET, IE improvement and day-to-
day management.
concluded my presentation with the statement that if
I I

were an American manager, would do my work by taking into

I

account three conditions and one difference (which follow)

and thus secure high quality and productivity in my opera-
tions.
Many missions and other
are sent from the United States
countries to study lapanese management. Our company,
Meidensha Electric Manufacturing Co. Ltd., welcomes more
and more of them every year Books and reports concerning
 Introduction xxii

Japanese management are published one after another.

It is often said that unique characteristics of Japanese
firms, such as lifetime employment, employee loyalty, the
promotion system, democratic decision-making, intra-com-
pany unions, etc, are the causes of successful management.
But if this were true, all lapanese firms would show high effi-
ciency and profitability.
In fact, there are many lapanese firms with these charac-

teristics which are also poor performers. Therefore, we must

compare not only Japanese and foreign firms, but also suc-
cessful and unsuccessful Japanese firms. The latter compari-
son will provide truly valuable information for those who want
to learn from successful Japanese management.
From my experience, 1 maintain that the managerial en-
gineering techniques presented in this book would have been
impossible without the following conditions:
( 1
Top management's support for managerial
engineering;
(2) The understanding and cooperation of labor unions;
(3) Employee interest and satisfaction in creative
activities.

As for the first condition, most of Japan's top managers

have long recognized the critical role of high product quality as
a source of competitive advantage, and expressed this need in
their management policies and leadership. They developed
company-wide quality goals and quality improvement pro-
grams, and made significant efforts to implement them. Quite
a few Japanese firms require their divisions to report quality
control activities to top management. At the same time, huge
investments were made for large-scale, company-wide QC
education programs. It was this long-term QC education that
spread concepts and methods of quality control in Japanese
firms.
Top management's support is also observed in other
xxiv MANAGERIAL ENGINEERING

fields. At Meidensha Electric, both the president and chairman

attended the employees' study circle for personal computers.

Through this experience, all the employees learned what top
management was aiming for.

The strong support from top management spreads down

through every management level. The "don't speak" rule for
OET activity, discussed in Chapter 6.1, would have been im-
possible without the plant manager's support. If the plant
manager had refused to support this activity, foremen would
have felt obliged to speak out to show off. And if the plant
manager had not allowed them to spend as many as 190 man-
hours a month for meetings, the plant would not have been
able to reduce the defect rate by 90%.
During my presentation, asked what American managers
1

do to encourage managerial engineering and what invest-

ments they make for long-term results, it is true that Japanese
firms do possess favorable characteristics that facilitate plant
management, but these are not sufficient conditions for suc-
cess. Actual efforts, such as those made by top management,
are much more important.
Incidentally, lapanese plant management is often de-
scribed overseas very selectively. One hears of employees
singing the company song and doing exercises every morning,
and even the manager's knowledge of the traditional Noh
dance that helps him make managerial decisions. Even the
lapanese people are unable to explain how these factors en-
courage high quality and productivity. This kind of misinfor-
mation is often brought to foreigners by the Japanese them-
selves. They will kindly explain what foreigners wish to hear
about Fujiyama and shogun. regret that this kindness brings
1

only misunderstanding and even suspicion on the lapanese.

To return to the subject, the second condition concerns
the cooperation of unions who work with management to
achieve high performance goals. The first goal of the union is
 Introduction xxv

to promote the employees' welfare and interests, but it can

also contribute to the prosperity of the corporation, which in

turn leads to employee satisfaction.

Mr. Donald F. Ephlin, vice president of the UAW, gave the
keynote address at the New York conference. He affirmed the
union's willingness to cooperate with the auto companies as
long as this is for reasonable goals. It would appear that this
second condition is becoming a reality in the United States.
The third condition expects workers to work not only for
money (which is the principle motivation in any country, of
course), but also for the pleasure and satisfaction obtained
through creative activities such as improving operations.
It seems to me that these three conditions now exist in
the United States. Emphasizing these three conditions, it is

my goal to provide the reader with correct, unbiased informa-

tion on lapanese management that will help them understand
their own situation.
Now, there an important difference between Japan and
is

the United States, once heard an anecdote: a group of

1

lapanese tourists was traveling in the United States. They or-

dered a rental bus at a hotel. The next day they got on the bus
and waited for the driver, who never showed up. The hotel
manager explained to them that they had ordered a bus, but
not a driver. They should have ordered both. This is an in-
teresting story which illustrates a fundamental difference be-
tween our two countries. "Doing only what is ordered" is inter-
preted as a criticism in lapan. But what if the hotel manager
above had been a computer, not a human being? The person
who input a vague instruction would be at fault, not the dull-
witted computer.
English has an expression, "butting in," which means in-
terfering in someone else's job or activity. In countries other
than lapan, this kind of butting in is strictly prohibited by cus-
tom, and the work system is designed so as to clearly delimit
xxvi MANAGERIAL ENGINEERING

each worker's area of responsibility.

This is the main difference between lapan and other
countries. When group provides ideas for an im-
a worker
provennent project, they are butting in on what is traditionally
the responsibility of engineering. But it is possible to develop
a work systenn where constructive interference of this sort is
not only accepted, but expected to contribute significantly to
higher performance.
If worker group activities are to be used in participative

management schemes in the U.S., this sort of behavior will

have to be encouraged. Since many American companies are

now developing QC circles, would think that they have started
I

dealing with this problem. sincerely hope that American

I

firms obtain the three conditions, overcome the difference,

and thus succeed in the development of new, American man-
agerial engineering.
On the other hand, Japanese managers must learn the
American way of work in the age of automation.
Computers cannot work in the lapanese system of reach-
ing out beyond one's limited area of personal responsibility.
Under Japanese management, the worker is constantly ex-
pected to reach beyond the strict limits of his job.
It is imperative for American management to start think-

ing along these lines. The lapanese, on the other hand, will
have to learn how to restrain this type of behavior. Automation
is only one reason for this. The other is that the younger gener-

ation in Japan is becoming increasingly westernized. It is obvi-

ous that in this world of international communication, lapan
will not be able to maintain its unique values and traditions

forever, as the law of increasing entropy suggests.

As American management
approaches lapanese
techniques, the latter must approach the former from the op-
posite direction. Once they are implemented, reliable
methods and practice are a permanent source of wealth for the
 Introduction xxvii

manufacturing firm, but the situation in wfiich this wealth is

used is subject to changes in the external environment. The
firm must predict future changes and take appropriate action.
Both American and Japanese firms according to their
will find,

characteristics and decisions, some appropriate balance be-

tween American and lapanese management.
it is important for Japanese firms to explore ways of de-
veloping QC circles in the United States. This will not only pay
off for American firms, but for lapanese companies ten years
down the road.
When had 1 finished this presentation, many participants
came up to shake my hand. At that time was impressed by theI

great strength of the American people: the will and ability to

think frankly and try to improve themselves. The audience
evaluations of my presentation, sent to me by Norman Bodek,
president of Productivity, inc., were very high. In fact, their
evaluations expressed more their impression of lapanese
management than the quality of my presentation. It was truly
encouraging to see that the ideas and approaches to manage-
rial engineering which have explained in this book were so
1

well received by these representatives of American manage-

ment. Furthermore, the conference provided the opportunity
to publish the English-language version of this book.
I wish much success to all managers as they set out to im-
prove productivity and quality, and hope this book will prove
useful to them in that quest.
Ryuii Fukuda
Figure 1-1 Driving Power for Management Activities
chapter 1

PRINCIPLES OF
MANAGERIAL
ENGINEERING

1.1 THE SEARCH FOR A RELIABLE METHOD

Is there a management method which can always reduce man-
ufacturing defects by 50% or yield a fixed rate of productivity
growth whenever it is used? "Too good to be true," the reader
may be thinking.
1 think that managerial engineering techniques such as
QC (Quality Control), IE (Industrial Engineering), and OR (Op-
erations Research) all originated in the demand for a reliable
An expert working alone will achieve very little with
method.
management techniques such as QC and IE. Working together,
however, a group of people can use these techniques to secure
significant results. My past experiences — including both suc-
cess and failure — have convinced me that when we as group
a
work towards the completion of a new project, we can only
push forward if we possess a reliable method. No matter how
often we talk to workers about improving product quality,
words alone will do nothing to promote real results. On the
contrary, we need to find a system, a method, which leads us
inevitably to success when its procedures are followed step by
step. As managers, we all want to obtain such a useful tool. 1

emphasize the word "tool." In fact, went through much study

1
 MANAGERIAL ENGINEERING

and research before concluding that managerial engineering

is a tool.
About two decades ago, Sumitomo Electric Industries
Ltd. the Deming prize. Prior to this event, the company
won '

had started an initial QC study program. At the time, was a 1

member of the Production Engineering Department. As was

usually the case with young engineers in the factory, was 1

overloaded with work. Then, out of the clear blue sky, head-
quarters created the new task of Quality Control. When I look
back on those days, cannot help but feel embarrassed. actu-
I I

ally thought my new QC responsibilities were unrelated to my

daily work. was completely unaware that QC was a useful tool.
I

Consequently, felt that my work had doubled because of this

I

new job. Aside from my urgent everyday tasks, now had QC to I

contend with.
Each department received lectures on QC from visiting
professors. As beginners, we took this new technique as the
golden rule. Therefore, when we were unable to understand or
apply this new knowledge, we tended to blame ourselves for
not studying hard enough. As a result, we dove into ever more
sophisticated QC techniques.
In 1975 was assigned to the head
I department office as IE
manager, responsible for managerial engineering, which cov-
ered QC, IE and OR. My goal was to help my staff and workers
avoid the mistakes had made in the past. Qver the years had
I I

developed my philosophy of managerial engineering: it is a

tool that can be applied by anyone to solve the problems of
even the most troubled section of a company.
To meet these requirements, a useful method had to be
developed from the basic elements of QC and IE. Since it was
developed from the basics, it was flexible and could be applied
'

lapan's prestigious award for quality, named for Dr W Edwards Deming. the
American statistician who introduced quality control to the Japanese in the
1950s
 Basic Principles

to a wide range of problems. However, to be useful this tool

had to meet another requirement, specialization, which is the
converse of flexibility. The tool also had to be tailored to the
company's specific needs.
Thus the basics remain unchanged, but the means of ap-
plying them change to suit time and place. A tool must serve
us. When we install a new piece of equipment, or use a newly
bought jig or tool, we improve it if it does not work well in our
plant. Likewise, since managerial engineering is a tool, it

should be adapted to meet our specific needs in the best pos-

sible way.
As a member of the research and development staff at the
beginning of my career, had many opportunities to work at a
I

variety of jobs. Before joining the IE Department, had been in I

charge of designing new products and developing new equip-

ment in the Technology Development Department. From my
experiences as the new IE manager, thought that the most
1

important thing in managerial engineering was to develop our

own methods, just as we did in other engineering disciplines.
We had to develop our own managerial systems, and
make sure that they would be more useful and widely applica-
ble in our company than the old methods. Developing these
ideas would also stimulate us to conceive new ones.
Study groups were organized to develop this new man-
agerial engineering. Small group activities involved employ-
ees from various levels of the corporate hierarchy and in-
cluded both white and blue collar workers. (The activities and
results of the first study group, known as the Quality Control
Problem Study Group, will be discussed in Chapter 3.)

I recommend that the reader develop his own methods of

managerial engineering by drawing on past experience and
the ideas of all of his employees. A managerial system de-

veloped by a given company should be designed in terms of its

specific business climate. This is especially true in the long
 MANAGERIAL ENGINEERING

run. If we seek only short-term results, we may find ways to

solve quality and productivity problems. However, we must re-
member that measures which focus on short-term results may
very well imperil the future of a company.

1 .2 CREATING A FAVORABLE ENVIRONMENT

Even if we possess a reliable method, we can expect favorable

results only if we establish the conditions necessary for its im-
plementation.
In 1980, I visited Sumitomo Electrics overseas plants in
Singapore, Thailand and Nigeria. Naturally, working condi-
tions in those plants differed greatly from what I knew in lapan.
In lanuary 1980, Sumitomo Electric started the Prosperity
Through Cooperation (PTC) program. It was a three-year proj-
ect, headed by the vice president, which aimed at setting up a
system for producing better quality goods at lower costs, with
faster delivery than our competitors. To reach this target, the
company directed its efforts towards improving communica-
tions and strengthening ties with its suppliers. The philosophy
behind the project was that both parties should benefit
through cooperation. As IE manager, was responsible for
I

promoting the PTC.

The most important part of this project was to communi-
cate its purpose to all the employees affected. We launched a
publicity campaign using posters, company magazines, etc.
We publicized the purpose of the program, its guiding princi-
ples and the important points that needed to be respected for
its implementation Meanwhile. began by visiting the top
1

management each subcontractor to explain our goals to

of
them. was astonished to learn that some of our departments
I

were working with companies located very far away. Some-

times when we visited those companies we were ushered into
 Basic Principles

the president's house adjacent to the factory. those cases,

In

their offices were too small to accommodate even our small

group. We spent much time on these visits because felt that 1

misunderstanding of our real purpose could sabotage our

goals.
The first thing understood when we began our tour was
1

that most of the subcontractors lacked an environment suita-

ble to work improvement activities. Some presidents showed
anxiety or doubt about our presence. One said, for example:
"You came to see our workshop. Please take a close look. Every
single worker is doing his job without resting. They work very
hard from morning to night. Where in this factory do you see
room for improvement? It is very difficult for us even to find the
spare time for any worker group activities."
Another said: "We employed a suggestion system as your
company recommended. Well, won't have anything more to
1

do with it. All that got from it was complaints about work and
1

demands for new equipment. lost interest in it. Before giving

1

the workers incentives to perform better jobs, prove to me that

this policy would improve productivity. In my factory, the
products are made by machines. am not interested in any
1

kind of management system."

Indeed, we saw many workers literally working by the
sweat of their brows under unshaded lamps. One said: "I, my-
went to several seminars on OC and IE. But have never
self, I

encountered a technique which meets our needs. thinkthose I

management techniques may be good for large firms, but not

for us."
These reactions were so extreme that we felt helpless.
While we listened to their opinions, spoken politely but with
candor, we clearly understood that the tough situation of the
suppliers would not allow us to introduce any managerial im-
provements unless we could make them appear exceptionally
attractive. We had to find a breakthrough. After seriously re-
 MANAGERIAL ENGINEERING

considering the matter, the managerial staff of our IE depart-

ment came up with a fresh idea that had never been used. We
would go to the workshops of each subcontractor and actually
demonstrate how to improve work right on the spot, in front of
the president and the other people in the company. The first
day we would observe how the work was being done. On the
following morning, we would do some of the work in our im-
proved way.
Company E in Osaka was the first subcontractor to be ap-
proached for this experiment. From among the many jobs we
observed, we chose simple ones to demonstrate our methods.
That night we met at the Sumitomo clubhouse and rehearsed
the jobs for the following day. was given the role of a table. My
1

duty was just to stand quite still, holding a tray neatly covered
with tools, alongside the other members who were going to act
as operators. This was a wise role in which to cast me; was too I

clumsy to handle the machine.

The next morning, the president, engineering manager,

15
Average Number of Improvements : 62
1 56%
Average Effect : 1 56%

Conventional Method (lime)

Improvement Effect = nm^5v ed Method (TirneT
" ^°°<°''°'

-. o> -i

Improvement Effect (%)

Figure 1-2 Results of Work Improvement

 Basic Principles

plant manager and other staff came to see our demonstration.

We did four jobs. The first was tool setting for the milling
machine. When we observed this task a day earlier, the
operator had taken thirty-seven minutes, even though he
worked hard under the added pressure of our cameras and
stop watches. To our great amazement, we were able to per-
form his task in thirteen minutes.
We decided to present a note of gratitude to each worker
who had allowed us to experiment on his job. The note said:
Dear Mr. :

Thank you for your cooperation in our experi-

ment. We were able to shorten the operating
time by 64% and lighten the labor by arranging
the tools properly, thus eliminating unneces-
sary motion. This is the basis of IE. Since you
have been doing this work every day, we are con-
fident that you will be able to improve it even
more than we did once you learn the basics of
IE.

There are infinite opportunities for improve-

ment, even in well established factories. Best
wishes for your health and work improvement.
Thank you,
IE Department
Sumitomo Electric

We found that the situation of the subcontractors im-

proved markedly after the demonstration. We continued to
visit the companies upon request. Partial results of the first

sixty-two visits are shown in Figure 1-2.

On the first people tended to say: "We

day of these visits,

have already tried the method you are talking about, but it
didn't work," or "We understand your point, but it is difficult for
us to do so." However, they stopped saying so on the following
 MANAGERIAL ENGINEERING

day, after witnessing our demonstration. Tiiey usually showed

strong interest nnethods and the basics of IE. One
in reliable

president later said: "The night of your visit, could not sleep,
I

thinking about what you had been able to do. The impact of
your demonstration on me was somewhat akin to Commodore
Perry's first visit to lapan in 1853, ending her two centuries of
isolation."
As shown in Figure 1-2, the effect of work improvement
varied from 1 to 3 1 0% with an average of 1 56%. The old "con-
ventional method "is the numerator of the formula, it gives the
operating time of the subcontractor's worker. The de-
nominator ("improved method") is the time the task required
using our methods. It is true that the workers in those com-
panies work very hard, but the results indicate that their oper-
ation time can be reduced by 56% if basic IE techniques are
used. The important point is that we increased productivity
56% by eliminating waste, not by accelerating operation
speed. With our experience and the data from this experiment,
we concluded that improvement must make work easier and
more efficient.
Another factor to be stressed is that a 56% reduction was
made against the value-added time. When only the incidental
work, which is less than 20% of the total work, is targeted to be
shortened, the workers will feel that making
the IE engineer is

their work harder by tightening up their schedule. Through

demonstrations, we were able to communicate to workers that
our intention was to ease everyone"s work and lower costs.
Seeing the acceptance of our demonstration for the sub-
contractors, we developed an IE course as part of the PTC
program. The course consisted of three eight-hour classes
which presented the basic techniques of IE, tailored to our
own needs. It was available to the employees both of our own
company and of our subcontractors In effect, we had dis-
covered a critical fact: education is one of the most important
 Basic Principles

Productivity Growth at Company E

250 -

^
H— 200 -

o
T5
O
150 -
O
CO

100 -

§ ,
April May June July Aug. Sept.
O) "

1981

Fiscal Year (April-March)

Figure 1-3 Productivity Growth at Company E

factors for creating a favorable environment.

Figures 1-2 and 1-3 show the growth of labor productivity
in Company E, the first company we visited. This company se-
cured a 40% improvement in productivity in less than a year
(froml980 to 1981). What is more, this was achieved at almost
no added expense. have introduced this example not to boast
I

about our success, but to show that anyone can do this. Any
company can establish an effective managerial system if it fol-
lows the basic principles of IE and develops a reliable method
suited to its own production conditions. The only other re-
quirement is practicing the method regularly.
1would like to emphasize that IE techniques are not
10 MANAGERIAL ENGINEERING

something only a few experts can use. IE was born to be used

by anyone, anywhere. It is a warm science, based upon human-
ity. It should not be used to threaten or embarrass anyone.

1 .3 PRACTICING TOGETHER

In Figure l-I, point C

"Keep all workers practiced in the
says;
method." For example, we were able to successfully dem-
onstrate jobs at the subcontractors' because we frequently
"got our hands dirty" on the shop floor of our own company.
Furthermore, we had read and heard many reports on im-
provement activities. Thus we had unconsciously practiced
floor level operations, and accumulated first-hand experience.
rarely play golf, but when talk to my friends about the
I 1

practice of management techniques, usually tease them by I

citing their practice in golfing. Where management techniques

are concerned, perhaps the word "training" is more common
than "practice." I prefer the word "practice," however, because
it implies a voluntary effort of will.

When you want to learn howtoplaygolf, you will probably

go out and buy book or take lessons. Then you hit a
a "how-to"
ball for the first time. If your shot is poor, what will you think?
Will you blame the book? Where golfing is concerned, people
usually do not blame the book for their poor shots. Instead,
they modestly regret their insufficient practice. After work or
on a weekend, they go to a driving range and practice.
On the other hand, people tend to spend little effort on
fully understanding a manual on management techniques, or
they abandon techniques that do not work on the first try. Then
they look for another manual. If the second one is not good,
they seek something new without practicing. With this sort of
attitude, how can they ever hope to perfect their management
techniques?
In fact, I once bought a book called Miracle Golfing. How-
 Basic Principles 1

ever, my golfing skills have not improved at all. That is simply

because I do not practice faithfully. The necessary conditions
for victory in managerial engineering are:
• Developing a reliable method
• Creating a favorable environment for the method
• Improving everyone's skill with it through constant
practice
Looking for powerful management techniques without
the benefit of a reliable method is like playing a game with
constantly changing rules. Winning such a game would be
nothing short of a miracle.
The proper amount of (A + B + C) in Figure 1-1 deter-
mines the degree of success in implementing a project
Whenever we start a new project, we can foresee our chance of
success by asking those three questions: Do we have a reliable
method? Is the environment favorable? Is everyone familiar
with the method? If the answers are negative, what right do we
have to expect success?
According to my experience, one of the most important
considerations is that (A -I- B -I- C) be a little bit greater than
required. When the group's driving power (A + B -t- C) is
greater than the difficulties that must be overcome, it can win.
divide management power into three factors because
I

this facilitates analyzing the situation. For instance, suppose

we have accomplished Project in six months. This means that
I

our present situation of A, B and C is better than it was six

months ago. We have made progress towards developing a re-
liable method. The environment has been getting better. We
have practiced the method repeatedly. What took a week the
first time now takes an hour. A setup that took hours can now

be performed in minutes. Since we have grown as a group, we

are now capable of attacking Project II, a more ambitious one
than our first project.
By following these steps, we will achieve the projected
12 MANAGERIAL ENGINEERING

goal, and at the same time, we will build our strength. There
may be strategies for attacking projects which require more
power than we have, and even if it is possible to achieve a
higher goal with extra effort, we may be exhausted. In view of
we achieve, the difference between two
the results strategies
may appear insignificant. But no matter how small it may be,
that difference is still there. Since success in the long run may
be measured by a narrow margin, it is always important to
maintain our competitive advantage.
chapter 2

RELIABLE WORK

2.1 LESSONS LEARNED FROM ACCIDENTS

In 1972, was working at the Facilities Developnnent Section of

1

Itami Works. It was divided into two sub-sections. Facilities

Planning and Testing was responsible for developing and test-

ing new equipnnent as well as planning and designing produc-
tion facilities. Its responsibilities covered everything from pur-
chasing materials to installation. The other sub-section was
the Maintenance Department, which was in charge of preven-
tive maintenance and repairs. It was broken down into two
groups: the electric maintenance group and the mechanical
maintenance group.
In the electric maintenance group there was a hydrogen

squad whose duties included checking hydrogen gas purity

upon receiving the gas from a tank truck, depressurizing and
supplying the gas to hydrogen gas tanks, and further depres-
surizing and supplying the gas to three plants and one labora-
tory. The squad had not had an accident since its foundation,
more than twenty years earlier.
in spite of this excellent record, I realized, as I watched the
television news report of a tragic lapan Railways disaster at the

13
14 MANAGERIAL ENGINEERING

Hokuriku Tunnel, that we had to improve the squads work

quality.
Two points made by the news commentator drew my at-
tention;
( 1 When the driver noticed that the train had caught on
fire,he stopped it in the middle of the tunnel. Why didn't he
drive out into the open air?
(2) There were about twenty employees on duty in the
train, including those who were working in the dining and mail
coaches. However, only a few of them actually tried to fight the
fire or lead the passengers to safety.
The first statement was obviously an afterthought of the
commentator. Common sense suggested bringing the train to
a full stop as soon as a fire broke out. The driver must not have
been taught that he should drive the train out of a tunnel and
then stop if a fire broke out. He had probably not been given
thechancetothinkaboutsuch an occurrence in hisdaily work.
And if such an accident had occurred with the hydrogen
at our plant... Had we taught the people in charge the neces-
sary emergency procedures?
Concerning the second point, wondered whether the
I

other people in the plant —

the maintenance workers and
users of hydrogen —
would be able to help the hydrogen
squad if an emergency were to occur. Probably not. We had not
trained those people to do so.
We had toconsider our firm's responsibility to society. No
matter how good our business performance was, if the hydro-
gen exploded, we might very well lose everything. (Unfortu-
nately, this thought was subsequently verified in the experi-
ence of another company.)
When arranged the order of our jobs according to the re-
I

quired reliability, realized that the highest priority should be

I

given to the prevention of a possible hydrogen disaster.

Although I strongly felt the need for emergency measures
and training, I didn't know where to begin.
 Reliable "^ork 15

2.2 DRIVEN BY NECESSITY

At that time, I was reading the Japanese translation of Loud and

Clear; The Full Answer to Aviation's Vital Question-. Are \ets Really Safe?.
a book by Robert Serling (Doubleday &Co., New York, 1969),
I.

which reviewed many airplane accidents. On the surface these

cases seemed unrelated to our work, but attempted to find a 1

common thread and glean some hints from it. tried hard to I

find a breakthrough, and did learn many lessons from the

1

book. The Appendix to this chapter gives some of them.

Based on these hints from Loud and Clear, and with the
cooperation of both the hydrogen squad and the plant's gas
users, we wrote a pamphlet entitled Emergency Procedures for Un-
foreseen Hydrogen Accidents. It listed all the possible unexpected

accidents we could imagine and suggested concrete coun-

termeasures for each case.
In May 1 pamphlet was distributed to all sections.
973, this
In it we described the necessity of preventing hydrogen disas-
ters.

We also provided emergency training. While conducting

this training, we made the very thought of disaster taboo, but
stressed that if a disaster should occur, we would somehow be
able to fight it. We tested each procedure, then trained our
people over and over again. In doing so, often encountered 1

situations which made methinkthat ifwe didn't do something

about the present state of affairs, we wouldn't be able to han-
dle a real emergency. was very much relieved that we had
I

started this training.

The next step was to organize a special squad responsible
for assisting thehydrogen squad in case of accidents. For the
special squad, five top persons were selected from other
maintenance groups. We began the education and training of
the squad. We trained five people to a level that would allow
them to work as efficiently as the members of the hydrogen
squad, were an accident to occur.
16 MANAGERIAL ENGINEERING

In the process of training the new nnembers, both the

nnennbers of the hydrogen squad and Irealized how differently
their work appeared when observed in detail. Annong the vari-
ous ways of doing the worl<, there must be one best one for pre-
venting hydrogen disasters.
The hydrogen squad spontaneously set about standardiz-
ing its work and eventually established the operations stan-
dards for twenty-four different jobs. As far as knew, this was I

the first voluntary activity of its kind. What had driven them to
create these standards and then train themselves to apply
them? Was it a strong desire to find the best possible way of
preventing accidents?
From these experiences, 1 learned the following:
(1) Exhortation alone is useless. "The duties of the hy-
drogen squad are important, so do your best." too, had often I,

subjected employees to such advice. When began this project I

to establish concrete prevention measures, realized for the 1

first time how useless these phrases are in practice.

(2) learned what work standards really meant Later on,

I

when started work as corporate IE manager, set up a stan-

1 1

dardization study group (the Quality Control Problem Study

Group).

2.3 EXTRACTING THE ESSENCE

In addition to these two lessons, I learned something even

more important. began to think in the abstract. The know-
1

how for preventing hydrogen disasters came from studying

airplane accidents. 1 think it is critical to be able to abstract the
essence from the bare facts
The word "abstract" is often viewed unfavorably in the
business world. People often say; "Don't speak in such an
In this case however,
abstract way; be specific." want to use 1

the word in the sense of extracting an essence. When we see or

hear something different from our daily work, we need to be
 Reliable \Mork 17

able to extract its essence as a possible hint for solving our

problenns. It will become more and more vital for us
businessmen to become skilled in extracting the essence.
When we are close to the top and still want to learn from
others, we have to be good at this skill. Furthermore, 1 believe
that the ability to extract an essence is crucial to the develop-
ment of new techniques in managerial engineering.
1 noticed an interesting phenomenon. When we as a
group visited another factory, shop-floor workers easily found
common points with their own work and learned from them,
even though the jobs looked quite different at first glance. On
the other hand, white collar workers tend only to look at the
different points. They say: "We are manufacturing the same
kind of product, but ours is smaller than theirs. That is differ-

ent.The functions of this machine are similiar to ours, but

from a professional viewpoint see there are differences be-
1

tween the two, cannot learn from it."

1

What is the difference between these two approaches?

Analysis and integration of facts are basic to IE. Both the blue
and white collar workers are observing analytically. The funda-
mental difference lies in whether they can unite the analyti-
cally observed individual fact with a common thread. believe 1

that finding these common threads is professional work in a

real sense.

2.4 TAKING A NEW LOOK AT HUMAN ERRORS

In 1973 a series of disasters, especially at petrochemical com-

plexes, occurred in Japan. It was reported in the press that the
causes of these disasters were human errors. gathered as
1

many detailed reports on the accidents as 1 could. Reviewing

the reports, I strongly feltthe necessity of preventive measures
for avoiding human errors. Of course, emergency measures are
important, but countermeasures against human error are im-
portant as well, since a human error in daily work might very
 MANAGERIAL ENGINEERING

well trigger a serious accident. This concern led us to develop

a new project.
We began by clarifying the potential dangers in our work
processes. Then we went on to find ways of preventing poten-
tial accidents. We made checklists according to each work pro-

cess. Our aim was to establish fail-safe systems for possibly

poorly designed operations or human errors. We reviewed all
points of motion and checked every step in operations. Table
2-1 is an example Looking back on it today, feel that it in- i

cludes several primitive and inadequate points, but it is the

table we used at that time.
While carrying out this project, we made more than thirty
improvements in tools, machines and operating methods. We
clearly perceived where we had to focus our training efforts IE
engineers, mechanical designers and architectural designers
participated in the project, and through our group effort, we
gradually learned what reliability engineering was all about. I

would like to stress that we learned from practice, driven by

necessity, not from textbooks.
Since had learned so much from Loud and Clear, con-
I I

tinued to read books about airplane accidents. Among my

readings were Mahfia No Kyofu {Panic at Macfi and Zoku Mahfia
1 )

No Kyofu (Aftertfie Panic at Mach by Kunio Yanagida, published

1 )

by Fujishuppansha. obtained much information from these

I

books. What conditions contribute to human error in a given

system? What are the major points we should analyze? The fol-
lowing are examples of what 1 learned:
Stress and tension (theory of errors)
Purpose of verifying meters
No two accidents occur in the same way
Redundant design function does as intended
Heinrich's principles (information flow for minor
accidents)
The use of fail-safe systems
Reliable V^ork 19
20 MANAGERIAL ENGINEERING

It is said that approximately 60% of all airplane accidents

are caused by human error. No matter how far technology ad-
vances, people are ultimately responsible for error-free opera-
tion. I am human beings, should take all re-
not saying that we,
sponsibility for accidents. mean that we should be fully aware
I

that people do make errors and that we must construct

machines and systems which take this fact into account. This
experience later led to the founding of our Human Error Study
Group.
While we were working aggressively on this project, an ac-
cident almost occurred. Mr. Umakichi Kurano of the hydrogen
squad detected an abnormal flow of gas and took speedy ac-
tion, thus preventing an accident which might have pulverized
the entire plant.
One day 1 read in the newspapers that Mr. Kozo Kitamura
had been studying gas explosions for the past twenty years. I

asked him to come check our works. "In most companies," he

commented, "more exciting work gets everyone's attention,
and the dull jobs such as receiving and supplying hydrogen are
relatively ignored. What makes you so active in this area?"
Whatthe abnormality detected by Mr. Kurano had taken
if

place before we had started our work? At that time, all could 1

give the workers was an announcement saying: "The hydrogen

squad has a great responsibility; do your best." When wonder I

whether he would have been able to take appropriate actions

under those circumstances, break into a cold sweat. cannot
I I

deny the relationship between what he did and the fact that
his team was trained for these eventualities.

2.5 THE DAY THE "GRAY WORKSHOP' CHANGED

Lastly,through our hydrogen-disaster prevention experience, 1

learned the critical importance of motivation. Towards the end

 Reliable "Work 21

of the project, I came upon the members of the hydrogen

squad making a model of the hydrogen distribution system, al-
though had not ordered them to do so. The model was so de-
I

vised that everyone could see at a glance the flow of gas, the
distribution of the pipes, and the function and location of the
various meters and valves.
"Why did you make this magnificent model?" asked I

them. They replied that in the event of an emergency, they

would immediately call the people in charge of the three
plants and the laboratory in front of the model. By using it,
they would be better able to command the situation.
By that time, special telephones for use in emergencies
had been installed. During the experiment it became evident
that the regular telephones were not enough, because there
were moments, especially during the night shift, when no one
was available near the phones.
was quite moved to witness this independent activity.
I

Were these the same people who used to go on in the

same old rut in the section nicknamed the "Gray Workshop"
because of its elderly workers? It was really an eye-opener. The
Facilities Development Section was a part of the Technology
Development Department. At the time the model was being
developed, was assistant manager of that department. We
1

were not necessarily lacking in lively projects. Development of

new products and facilities and construction of a new plant
were also progressing. However, believed that the prevention
1

of potential hydrogen disasters was our most important

mission.
This was not a reflection on the other projects. simply 1

felt that were an explosion to occur, it might result in the total

disruption of our activities.
From this project I learned the essential conditions for
motivation. Whatever the job may be, when people recognize
its necessity and importance, and when the required measures
22 MANAGERIAL ENGINEERING

are taken, they become highly motivated. This is a fact, not a

theory.
In a good plant or shop, people should feel as excited as
Paul Revere dashing through the night with his passion for rev-
olution over two hundred years ago. What should I call it? Mis-
sion? Aspiration? I believe that the ultimate factor supporting
quality work lies in each individual's mind: aspiration —
something spontaneous.

APPENDIX TO CHAPTER 2
CONVERSION OF LESSONS FROM AIRCRAFT ACCIDENTS
TO PREVENTIVE MEASURES AGAINST HYDROGEN
ACCIDENTS •

On December 20, 1966, more than a year after the collision, the
Civil Aeronautics Board issued a report which found an optical
illusion responsible for the mishap. The CAB said the cloud
bank just under the Constellation in reality sloped downward,
giving White and Holt a false horizon. They actually were sepa-
rated horizontally from the jet by the required one thousand
feet — Eastern had been assigned to ten thousand feet and
TWA to eleven thousand — but were fooled into thinking that
TWA was at their altitude. Likewise, the TWA pilots also took
evasion action because of an identical illusion — they assumed
EAL 853 was at eleven thousand. The supreme irony was that
the alertness of both crews triggered the collision. It would not
have occurred if one or both had not spotted the other, (p. 68)

Altimeter trouble, either a malfunction of a misreading, could

have been a factor in this accident and it may have played un-
suspected roles in other crashes blamed on everything but the
altimeter, (p. 140)

'
Excerpts from Loud and Clear, see above, p. 15.
 Reliable "Work 23

(Are our meters always reliable? If not, a person will instinc-

tively rely on his judgment at the last moment.)

The NASA report to the CAB pretty well pinpointed the real
reason for at least three of the four tragedies — failure in airline

training programs to emphasize the virtues of the 727 as a po-

tential hazard. Namely, that its ability to make short landings

could also be a booby trap for any pilot who does not fly the 727
by the book. NASA informed the CAB that close-in, so-called
"unstabilized" approaches with high rates of descent, were
being conducted more often in the 727 than in any other jet

transport, (p. 178)

(Likewise, we must consider the safety of new facilities with re-

gard to hydrogen from the initial designing stage.)

The second lesson, however, was the direct result of the Salt
Lake City accident — the tardy realization that passengers need
better odds for getting out of a theoretically survivable crash.
(p. 179)

(In plants, too, there must be ways to prevent a disaster result-

ing from a minor accident.)

"...the captain's voice came over the intercom and in a very re-
assuring manner he told us that if we would obey orders there
was no need to panic; that everyone should put on his life jack-

et; take off shoes; place his pillow on his lap; and lean for-
ward...He also told us to remove any sharp obstacles Ihe
meant objectsj and wrap our hands around our ankles." p. 58) (

(In case of hydrogen accident, who is going to be a command-

er? Is that person trained?)
24 MANAGERIAL ENGINEERING

The cabin was quiet: the six flight attendants — the purser
and five stewardesses — had everything under control and
ready for landing, (pp. 58-59)

"As one of those one hundred and seventy-four guinea pigs, 1

unfastened my belt and started stumbling through the smoke

toward the rear — where I remembered boarding. (My reaction
was typical of the majority of all airline passengers: I instinc-
tively headed for the door through which boarded, even though
I

"
there was an emergency exit much closer.) (p. 181)

(We have to get nnore information on human behavior in an

emergency.)

C. Hayden LeRoy, an investigator for the CAB's Bureau of

Safety, later did a thorough and valuable analysis of the evacua-
tion flow, as well as other survival aspects of the UAL accident.
His study provided an interesting correlation between the seat
locations of surviving passengers and the exits they used. The
majority followed an escape route toward the entrances used
when boarding. some cases, this contributed to the fatality
In

toll, because a number of passengers were overcome by smoke

and flames before they could reach exits at the opposite end of
the plane, (p. 184)

Obviously, there is considerable "it always happens to the

other guy but it could never happen to me" thinking on the part
of the public or, for that matter, even of the part of some crew
members. This ostrich attitude could be placed in the same
category as smoking three packs of cigarettes daily or driving a
car at sixty MPH after downing five martinis. It is a foolhardy at-

titude because no one can predict an accident and no one, even

more significantly, can predict his own reactions if an accident
occurs, (p. 184)
 Reliable \S/ork 25

(Don't think that we can predictanaccident. What we can think

of is only a part of it. In reality, an accident attacks us suddenly
in a way we could never predict.)

"You know why you're here," Clancie Imanagerof steward-

all

essesl began. "To make sure you know what to do in a few min-
utes or even seconds to assure the safety of the greatest number
of passengers possible. And mean seconds. If something goes
1

wrong with no warning, you won't have time to look up proce-

dures in your manual." (p. 201)

"All of you admitted you've never tried saying it aloud, in prac-

tice. Lock yourself in your bathroom if your roommate laughs at
you. Don't trust to luck any phase of emergency procedures. The
chances are that not a singlegirl of this room or in this base will

ever have tomake that PA. But this doesn't mean you shouldn't
be prepared to make it." (p. 204)

Now the class proceeded to unplanned emergencies and the

questions came at the girls like machine-gun bullets.

How do you handle an incapacitated passenger in a crash

landing?... What do you do in case of an in-flight cabin fire?...

Where are the emergency stations for the first stewardess. ..sec-
ond. ..third?...
Then came the written test. Questions like:

During your briefing for a planned belly landing in a 727,

where would you tell a passenger sitting in 19A to try to

evacuate first? " (p. 204)

(Sinnilar questions might be: You are on night shift, working at

a hydrogen furnace. The south end furnace of the line has just
gone up in a blast, the next one catches fronn the blast, and it
seems as though explosions will continue in a line toward you.
What would you do at the moment? Or: While you are filling
26 MANAGERIAL ENGINEERING

the hydrogen gas into a gas holder from a tank truck, the con-
necting pipe comes off, and the gas is blowing off violently. It

hasn't caught fire yet. What do you have to think of first? What
actions would you take immediately?)
 chapter 3

A RELIABLE METHOD ( 1 ):

THE CAUSE AND EFFECT

DIAGRAM WITH THE
ADDITION OF CARDS

Improving product quality, decreasing variations in quality

and reducing the defect rate are basic and essential tasks for
manufacturing firms. In this age of low economic growth and
keen competition, superior product quality has become a key
source of competitive advantage. Since improving quality and
reducing defects are essentially one and the same the dif- —
ference between the two lying in where one chooses to set
quality standards and limits — we will approach the problem
from the latter point of view for the sake of simplicity.
The term quality control, and the concepts it implies, are
widely known throughout lapan. A few years ago, was visiting
I

a factory in the countryside on a hot summer day. On the way

I stopped at a small restaurant and ordered iced noodles.
When a young waitress appeared with the order, commented I

that one of the bowls was cracked. She replied immediately:

"QC for this bowl wasn't very good!" — a remarkable indica-
tion of how popularized QC has become in lapan.

27
28 MANAGERIAL ENGINEERING

3.1 IDEAL CONDITIONS

About seven years ago, we were looking for an effective

method for reducing our defect rate. We were especially con-
cerned with two kinds of defects: those for which we were un-
able to identify the cause, and those which were caused by the
our processes to achieve designed quality levels.
inability of
These problems often appear in the early production stage of
new products.
When such defects appear, people often blame weaknes-
ses in operation and engineering standards for the product.
Since we did not yet know all of the causes of these defects, it
was impossible for us to write truly effective and practical
standards. First of all, we had to establish standard production
techniques and operations. Whether or how they were to be
described on paper was a secondary problem. We set our
sights on creating an easy method that would allow everyone
to develop and use standard techniques and operations.
We had also been annoyed by another QC standardiza-
tion problem. The more our production process produced de-
fects, the more we needed improvements in operations,
manufacturing conditions and equipment. The revised stan-
dards never caught up with frequent improvements made in
processes. Since no factory had excess manpower to assign to
the task of revision, the standards remained perpetually out of
date. Consequently, the differences between actual opera-
tions and official standards led workers to ignore the latter.
Timely revision of standards was also difficult in the fol-
lowing situation. When we effected an improvement in manu-
facturing equipment or conditions, we usually had to wait
some time before we actually observed a decrease in defects It

was virtually impossible for us to update standards in our daily

work after each of these improvement cycles. In many cases,
no one bothered to revise them. We wanted a system by which
 Reliable Method: CEDAC 29

truly reliable standards were constantly provided and adhered

to by everyone.
Table summarizes the above situation. It also de-
3-1

scribes the ideal conditions to be realized from an imperfect

production environment.

3.2 THE QUALITY CONTROL PROBLEM STUDY

—
GROUP CONDITIONS FOR DEVELOPMENT
The Quality Control Problem Study Group was formed in
November 1975 under the direction of the Industrial Engineer-
ing Department. Its goal was to develop a method suitable for
securing the ideal conditions given in Table 3-1.1 was ap-
pointed manager of this department in luly of that year.
The group originally consisted of thirty-two volunteers
with various positions in the company —
plant managers
(Sumitomo Electric has nearly forty plants), foremen, produc-

PRESENT CONDITIONS
30 MANAGERIAL ENGINEERING

tion engineers, industrial engineers, etc. It included members

from the ranks of management, staff and blue collar employ-
ees, in other words, it was a small, vertically integrated circle. 1

felt that such a formation was desirable for the development of

a new method.
The group was set up along the following principles:

(1) Although we were working towards a single goal [ideal condi-

tions), the thirty-two members were instructed to develop thirty-two dis-

tinct approaches.
When faced with a difficult problem, the Japanese tend to
stickunswervingly to whatever idea they happen to hit upon.
By thus manifesting their determination, they expect to win
the sympathy and respect of others. felt that it would be much
!

more useful to have each member work creatively on his own

individual proposals. Thus the group would benefit from a
variety of ideas.
(2) Each member was to work on a problem drawn directly from his

daily work in the plant, for which he would develop and test his own method.
I asked each member to do this as though he were "nursing a young plant."

Picking a problem from one's own work was simple but

Choosing a problem that was very important for the
critical.

company, but did not come from his own work, could not work.
The main reason for this has to do with the member's motiva-
tion; hoped the members would develop enough zeal to at-
I

tend the meetings assiduously. If attendance did help mem-

bers with their own concrete problems, they would become
powerfully motivated. If, on the other hand, they were to deem
managerial engineering a further burden to their already
crowded schedules, they would give up easily. Managerial en-
gineering must be a useful tool that helps us in our day-to-day
work.
used the above image of "nursing a young plant" be-
I

cause each group member was required to bring the same

methods to problem solving as one uses to nurse a young
 Reliable Method: CED AC 31

plant — knowledge and experience. Furthernnore, once a

month all the mennbers brought their "plants" to the meetings
to show to the others. By observing them and sharing know-
how, each member could learn more. Our intention was to
create a new method by integrating all this knowledge and ex-
perience. Even though each method in itself often contained
nothing new, we created a new approach by integrating all
these diverse methods and knowledge into a single, coherent
approach.
Looking back on the beginning of the study group, 1 sup-
pose the members may have had many doubts on their minds,
such as: "The only reason the company called this group to-
gether is to impose standardization measures," or "Perhaps
the new IE manager is going to tell us about his new ideas and
then force us to use them."
Some them may have attended the first meetings with
of
resignation. Theymay have thought: "We have already tried
many different methods and none of them have met our ex-
pectations. Our problems are really tough. Perhaps we
shouldn't expect too much from this meeting, either."
Other members may have had ready-made explanations
for their not being able to achieve ideal conditions. made a 1

point of explaining to members that since existing methods

had proved inadequate, our goal would be to create new
techniques improving quality and productivity.
for
It is difficult to start something truly new. In fact,
always
we were quite anxious about how the group would turn out. An
IE staff member confessed to me his initial feeling — that if

the group was able to meet more than three times, that would
already indicate a certain degree of success.
(3) Meetings were held monthly at each of our five works, and the

number of members attending each meeting was generally not more than
ten.

The members functioned as liaisons between the group

32 MANAGERIAL ENGINEERING

and the plants and also led QC activities in their respective

plants. In this way we learned much about leadership, and we
found that the positions of the members did not necessarily
represent their levels of leadership activities in their plants.

(4) The groups operated on the principle of open membership. This

meant that anyone interested in a group could enter it freely, and anyone
who was dissatisfied could leave.

The number of people attending the meetings grew to al-

most 100. This number was not constant, however, since indi-
viduals were free to join or leave as they wished.
When I explained this study group approach to managers
from other companies, they often expressed astonishment at
our management's generosity in allowing the group members
to devote company time to such a dubious enterprise. would 1

like to stress that it was this warm support both from manage-

ment and from the line that enabled us to bring our project to
fruition.
The study group developed a technique for improving
quality and productivity —
the Cause-and-Effect Diagram
with the Addition of Cards (CEDAC). Through this example, I

hope the reader will understand my way of thinking and

methodology for managerial engineering development.
This book is not meant to be used as a textbook of QC and
IE.Furthermore, rather than bore the reader with abstract dis-
cussions, cite actual cases, drawn from our experience at
1

Sumitomo Electric and Meidensha Electric Mfg. Co., to illus-

trate theoretical ideas. My goal is to leave the reader with a
sufficiently clearunderstanding that he may apply the princi-
ples to other situations. Even those managers who are not di-
rectly involved with QC or IE in production, such as marketing
or personnel professionals, will thus understand not only the
cases themselves, but their essence, and will be able to apply
this understanding to the quality and productivity problems
encountered in their own departments
 Reliable Method. CEDAC 33

3.3 WHAT IS CEDAC?

CEDAC grew out of the diverse approaches of the individual
group members. It is a modification of the cause-and-effect
diagram, which was already well known and widely used by QC
circles in our plants when 1 started our quality improvement
project.
An example of the cause-and-effect diagram is given in

Figure 3-1.
All the factors considered to be causes of a given effect
are written out in "fish-bone" form. CEDAC grew out of this dia-
gram. It emphasizes the importance of both the engineering

knowledge and workers' practical experience that lie behind

simple words such as "temperature" or "dryness" found on the
cause-and-effect diagram.
However, the traditional cause-and-effect diagram uses
single words, written at the appropriate place on the diagram,
to indicate the causes of a given problem and their inter-
relationships.
In order to solve a serious quality control problem, we
must start at the very beginning, distinguishing what we know
from what we do not know. For this purpose, a short sentence
is more desirable than a single word to fully express what is

known.
The use of small cards another point. In traditional QC
is

circle applications of cause-and-effect diagrams, the employ-

ees must wait until the weekly meeting to put their ideas on
the chart. If people have to wait before they can get together to
share information, they will often miss a timely opportunity to
solve problems. It is usually very difficult to gather together all
the workers on the shop floor.
Furthermore, people often feel very timid about writing
their own ideas directly onto a wall diagram. If they do write,
they tend to feel that they must express their ideas in a formal
34 MANAGERIAL ENGINEERING

Q.
E
I-
 Reliable Method. CEDAC 35

way. The use of cards solves both of these problems at one

stroke. Each worker is free to jot down an idea whenever it en-
ters his nnind.
The use of cards also facilitates revision of the diagrann,
since a new card can be pinned on top of the corresponding
old one. The succession of cards at any given point on the
chart may thus be interpreted as a history of changes occurring
at a specific point in the production process.
Most statistical quality control methods integrate quan-
titative information, but there is also a lot of important qual-
itative information available in the plant which is either non-
quantihable or not yet quantified. CEDAC is a tool that allows
the quality circle to integrate this kind of information. Raw in-
formation is indispensable for problem solving not only in
manufacturing, but also in marketing, personnel and in other

fields.Such information often evaporates quickly, leaving only

a minor numerical trace.
Once when was learning about a control chart at a QC
I

seminar, the instructor gave us an assignment to draw a con-

trol chart using numerical data in our own plants. In fact, I

found it very difficult because there was little numerical data

available in our plant which could be meaningfully written on
a control chart.
was my hope that CEDAC could be used to gather to-
it

gether the ideas of many people by comparing causes and ef-

fects, especially in those cases where qualitative information
plays a critical role. CEDAC encourages the exploration of
both favorable and unfavorable causes and looks for the ef-
fects and relations between them by permitting the systematic
analysis of facts.
The following paragraphs describe the application of
CEDAC to the manufacturing process. The basic principles and
techniques of CEDAC remain the same in other business ap-
plications. The reader is encouraged to interpret the rest of
this book according to his own needs.
36 MANAGERIAL ENGINEERING

3.4 STEPS FOR IMPLEMENTING THE CEDAC

SYSTEM

Figure 3-2 shows the basic structure of CEDAC. These five

steps are used to implement the CEDAC system:
Select a major quality problem that you want to solve
( 1

and specify the goal to be achieved, it is important to specify

the quality quantitatively, although this is not always easy.
(2) Write out all technical know-how and manufacturing
conditions that are thought to have an influence on quality.
From this, a selection of the needed information is made by
workers, production engineers, plant engineers and all other
concerned persons. Put all of this information in diagram
form.
(3) Hang the diagram on a wall of the plant for all to see.
This way, the causes and effect to be studied will be visible to
everyone.
(4) When quality cannot be kept within control limits,

One short sentence on a card descnbes

the necessary conditions and technical
know-how for controlling the specific
quality identified at nqht. One specific quality of the product.

w
 Reliable Method: CEDAC 37

room for improvement in the causes is indicated. Thus, the

root of the problem is sought through the gathering of more
facts, if necessary, the group tests its tentative countermeas-

ure for the problem, then observes and analyzes the results.
(5) Based on the analysis of Step (4), the group will de-
cide to make improvements either in technique or equipment.
The exact nature of this improvement is written down on a card
which is then pinned on top of the corresponding old card. As
a group of cards is gradually accumulated, it will show not only
the past record of the production process, but also the effects
of each improvement.
Everyone participates in the process from Steps (2) to (5).
With such a method, previously unknown information is dis-
covered, and necessary improvements are made on a step by
step basis, without any regression.
In Step ( a major quality problem must be selected. It
1 ),

has been my experience that everyone concerned must do his

best in order to create anything really new. Groups using
CEDAC for the first time often make the mistake of choosing
an easy rather than a difficult problem. This leads to a very
high risk of failure, because the group members can only ob-
tain insignificant inspiration and results by tackling a minor
problem, it is essential to tackle the most serious problem that
everyone is eager to solve.
The cause-and-effect diagram was used as the basis of
CEDAC. This is not only because it was already being widely
used throughout the company. More importantly, cause-and-
effect diagrams became the basis of CEDAC because it is es-
sential to analyze quality control problems in terms of cause
and effect.
However, we decided not to stick to the fish-bone form of
diagram. With the permission of Professor Kaoru Ishikawa, in-
ventor of the cause-and-effect diagram, we were able to use
any diagram as long as it clearly expressed causes and effects.
38 MANAGERIAL ENGINEERING

Regarding Step (2), Figure 3-3 shows the process of incor-

porating various bits of data. First of all, cards are collected
from all the group members.

The cards will often show a variety of approaches. When

an existing operational standard is not sufficiently good to
prevent defects from appearing, workers will not necessarily
observe it; they may try different approaches in order to de-
crease the number of defects. It is not important to know how
these standards are described on paper, in fact, it is not even
necessary to describe these standards on paper at all. The im-
portant issue is whether procedures or standards effective
enough to produce the desired quality are known or not.
 Reliable Method. CEDAC 39

For instance, Operator A happened to experinaent with a

new method during a rainy night shift. He observed that this
method produced zero defects and has adhered to it ever
he told it to Operator B. Unfortunately,
since. Quite naturally,
Operator B missed a few important details when he attempted
to use the same method himself. When they wrote their re-
spective methods down on cards, it became clear to both
workers that they had overlooked differences on critical
points.
Another example is the production engineer who
typical
discovered that workers were following methods completely
different from those that he had established forthem. This dis-
covery led him to revise his current experiments and plans to
take into account these discrepancies.
Thus, in many ways, management and engineering are in-

evitably confronted by the great variety in workers' methods

and know-how. At Sumitomo Electric we observed that this
variability often causes serious quality control problems. The
great power of the CEDAC method lies in its ability to integrate
all of the information concerning this variability on a single
card through fact-finding, OC circle meetings and experiments
(Figure 3-3).
An important point concerning Step (4) is that looking for
a cause after a given abnormality has been spotted by a con-
trol chart is not as effective or practical as observing effects by
making changes in the causes. Note that these changes may
be made intentionally or unintentionally, but always con-
sciously.
The approach brings more accurate, timely and ap-
latter
propriate information with less time and effort. It also enables
us to obtain information on what factors affect quality and
what factors do not change it at all. Moreover, it is a much
more interesting and stimulating approach, which is an impor-
tant factor in our daily work.
40 MANAGERIAL ENGINEERING

[Cause]

1
1
1 1
1
1

LU
CO

<
o
Reliable Method. CEDAC 41

o
42 MANAGERIAL ENGINEERING

W.A. Shewhart designed a control chart that signals an

alarm whenever a specific quality is outside the control limit,
at which point one is expected to determine the cause and
take appropriate action.
in a real plant situation, though, each worker follows his
own distinct work methods, and it is doubtful if we can really
use this method to discover true causes. have never suc-
I

ceeded in this.

Furthermore, Shewart's approach forces us to look for in-

formation. This difference between the two approaches may at

first appear trivial, as both aim at quality control through the
use of a control chart, but they have far different effects on the
participants of quality control activities. People prefer a posi-
tive state of mind — encouraged by looking for ways to im-
prove quality — to having constantly to be on the watch for
defects. CEDAC facilitates this approach and encourages a
positive attitude.
Of all the approaches to QC with which we have ex-
perimented so far, CEDAC has by far gained the most en-
thusiastic reception from our workers. One of the major
reasons is that it is an easy-to-use method for accumulating
meaningful data; workers use it voluntarily because it helps
them conceptualize correctly the production process, and thus
develop powerful suggestions for improvements.

3.5 SOME VARIATIONS ON CEDAC

In addition to the fish-bone, we devised many other types of
diagrams. Some examples are shown in Figure 3-4.
 Reliable Method. CEDAC 43

3.6 THE CONCEPTUAL FRAMEWORK OF CEDAC

As shown in Figure 3-4, CEDAC has taken on a great variety of
forms and functions developed by members of the Quality
Control Problem Study Group to meet their respective needs
and backgrounds. However, all of these variations share a
common conceptual framework, which is shown in Figure 3-5.
As (i) of Figure 3-5 indicates, the workers group is in the
most advantageous position for fact-finding. This is so
because:
• They are constantly observing raw facts
• Their sheer number helps in fact-finding
• Their accumulated years of experience have led them to
acquire tremendous insight into the production process

IMPROVEMENT AND MAINTENANCE OF QUALITY

BY PARTICIPATION OF ALL PERSONS CONCERNED

44 MANAGERIAL ENGINEERING

• They get their hands dirty in their work, and have a great
desire to solve the problems they encounter on a day-to-
day basis
People tend to underestimate the basic fact-finding func-
tion of all QC circle activity. CEDAC highlights this function,
which is one of the great strengths of workers, and aims at the
systematic integration of facts observed by individual workers.
On the other hand, production engineers possess en-
gineering knowledge that can help them discover the causes of
defects. All too often today, factual information provided by
workers is allowed to evaporate into mere tables of numerical
data. This information must be used by engineers to signifi-
cantly improve operations by discovering unknown factors
and developing improvement ideas.
The use of CEDAC, if adequately practiced, facilitates the
solution of the many problems which crop up in the constantly
changing production process. Daily practice is as important
for QC as for football. Furthermore, once a good method is
found, it should be retained. As have already stated, using dif-
1

ferent, ad hoc approaches to product quality is just like playing

a game with changing rules.

3.7 UNEXPECTED RESULTS

Our many plants thus began to use CEDAC for serious quality
problems with enthusiastic support from both management
and workers, and achieved truly remarkable results, such as
completely eliminating defects, or reducing them by 90% or
more.
Table 3-2 summarizes the results obtained in two of our
plants where all of the workers groups tried the CEDAC ap-
proach on a total of thirty-one QC problems The results were
obtained within periods of four days to eight months. In about
 Reliable Method. CEDAC 45

^\^,^^ Results

Problems ^^^^^
46 MANAGERIAL ENGINEERING

3.8 LESSONS FROM THE HAWTHORNE STUDY

To analyze the effectiveness of CEDAC, we applied research

principles learned from the fannous study which was made
about fifty years ago at the Hawthorne Works of Western Elec-
tric Company, Chicago, Illinois. With the development of

statistical methods and the computer, our analytical

techniques are more sophisticated today. What we learned
far

from this study was not the techniques, but rather a basic re-
search attitude necessary for managerial engineering.
Putting aside our CEDAC story for a while, a short expla-
nation of the illumination experiments which were part of the
Hawthorne study is in order. From 1924 to 1933 Western Elec-
tricconducted research and experiments on the factors that
influence worker morale and productivity. The illumination ex-
periments were the first part of the Hawthorne study. They
were conducted by management in conjunction with the Na-
tional Research Committee of the National Academy of Sci-
ences The study was led by Elton Mayo of Harvard University
and others, it is said that the study was so extensive that it pro-
duced several tons of documents.
The goal of these experiments was to measure the re-
lationship between lighting and productivity. In the experi-
ments five women were selected from many female workers of
the relay assembly room and transferred to an experiment
room. It was originally expected that decreasing the level of il-
lumination would decrease productivity. However, the results
were not all that simple. Contrary to expectations, productivity
increased. Productivity only decreased when the room was too
dark to see anything at all The experiment itself was consid-
ered a failure.
However, it is Hawthorne experimenters' credit
to the
that they perceived the existence of unknown factors behind
this unexpected increase in the women's productivity in spite
 Reliable Method: CEDAC 47

of the fact that they were working in a darkened room. It has

been said that this inquisitive attitude has opened up new
frontiers in engineering, such as nnotivation theory, organiza-
tional behavior, etc. And it is this attitude that we should
adopt in managerial science whenever we hit upon an unex-
pected phenomenon: wondering about it, looking for causes
and analyzing its mechanism.
The Hawthorne study was published as a book which in-
troduces the five young women complete with pictures and
very detailed biographical data such as the national origins of
their parents. Even the menus of their breakfasts before each
experiment were recorded. From the wealth of such data it be-
comes clear that the researchers were trying to approach the
problem as scientifically as possible.'
We decided to apply this attitude. By finding a scientific
explanation to our own unexpected discovery the far-reach- —
ing effects of CEDAC —
we were able to develop a reliable
method that can be applied by anyone, any time, anywhere.

3.9 THE HYPOTHESIS OF NEW JOHARRY'S WINDOW

Returning to the remarkable effectiveness of CEDAC (Table 3-
2), here we present a hypothesis we formulated to explain it.

As Figure 3-6 shows, loharry's Window was used by

Joseph Ruft and Harry Ingram to describe communication be-
tween two persons ("you" and "I"):
• Category I refers to what both you and know 1

• Category II refers to what know and you do not know

I

It became clear to us that the statistical analysis of the Hawthorne study may
contain certain errors; this point will be discussed in Chapter 8
48 MANAGERIAL ENGINEERING

^^^^/OU
Reliable Method. CEDAC 49
50 MANAGERIAL ENGINEERING

the standard operations, but there is someone who does not

practice them correctly. This includes the case where someone
fails to adhere to standard operations out of carelessness.
(7) In Category 111, one party knows but the other party
does not know the right operations for preventing defects.

(8) In Category IV, no one in either party knows the right

techniques yet. The technical problems which cause defects
remain unsolved in this category.

Next, let us consider the functions of CEDAC by referring

to the New loharry's Window (Figure 3-7).
Through the use of CEDAC, many actions were taken to
reduce defects (see Section 3.4). All these actions can be
explained as the efforts to transfer Categories II, III and IV to
Category That is, "communication of material in Category III,"
I.

then "adherence to established standard operations" and "de-

velopment of better production methods and technology."
After examining cases where CEDAC had been applied, we
found that the first two actions were in themselves enough to
achieve results. When quality problems remained unsolved
after those actions, though, the third action —
improvement
of technique and equipment —
was necessary.
However, adhering to the standard operation does not
mean following it without any deviation whatsoever. In fact,
deviating from present standards to test new methods by trial
and error is essential for development of better production
methods and technology.
To solve quality control problems, we must follow the
cycle Steps 1, 2, 3, 4, 1, etc. If any single step is overlooked,
good results may be compromised. The key to success is not
how to use CEDAC, but how to follow this cycle using CEDAC.
In the past, most quality improvement activities have
been a sort of search into Category IV —
the improvement of
technique and equipment At times, however, they have been
made without taking any action in regard to Categories II and
HI. Thus, the unstable and undefined elements of Categories II
 Reliable Method CEDAC 51

and III resulted in an incorrect decision by engineers, or in not

achieving results even when
good decision had been made.
a
Before implennenting CEDAC, we had often observed
cases where a new production level led to an initial reduction
in defects that was soon followed by a return to the original

level, in addition to poor comnaunication, other factors such

as carelessness entered into play to cause defects as time
passed. Even if some factor is originally part of Category 1,
there is no guarantee that it will remain there indefinitely.
Continual communications and efforts are needed to main-
tain Category CEDAC assures this communication. The dia-
I.

gram is looked at by all persons concerned, and cards with new

information are regularly placed over old ones each time an
improvement is made.
Our method for transferring conditions in Categories II, ill
and IV to Category could have important implications for
I

fields other than quality control. In a communications/infor-

mation-oriented society, where knowledge and informiation
play a key role, effective methods for perfecting channels of
communication will be at a premium. Management in this so-
ciety will have to provide a system in which all employees con-
cerned with a given problem share necessary information and
voluntarily participate in achieving shared objectives.

3.10 DEFENSE (ADHERING TO ESTABLISHED

STANDARDS) AND OFFENSE (FINDING NEW
METHODS)
CEDAC had now spread throughout the company, and the
number diagrams which had been drawn by April, 1978 al-
of
ready exceeded 350. To test the hypothesis of the New
loharry's Window, eighty-six cases were randomly selected
from twenty plants in ten divisions. The data from these
eighty-six cases were analyzed quantitatively, then synthe-
sized into a form easily understood by all employees. For em-
52 MANAGERIAL ENGINEERING

ployees to start practicing a new method, an understanding of

that method based upon scientifically verified data is the most
effective motivation.
The hypothesis of the New loharry's Window implies two
approaches to the solution of quality control problems:
• Devoting effort to Category (adherence to established
1

standards)
• Looking for new production methods in Category IV. Our
Our hypothesis is that the most productive route is to first
work on transferring presently known methods and know-how
to Category 1; if the problem is still unsolved, then it is time to
move into Category IV and look for new techniques.
Figure 3-8 classified the eighty-six cases into two groups:
(a) Groups which focused their efforts only on adhering to
established standards.
(b) Groups which succeeded in finding new operation
(production) methods;
The results shown here were tentative, and the possibility
of achieving further results was not investigated.
The reduction of defects by group (b) was attributed to
their adherence to previously formed methods. Their percent-
age of reduction is less than that of group (a). Therefore, of the
two approaches, it became clear to us that, for reducing de-
fects, discovering new methods is more effective than just

adhering to established standards.

An interesting conclusion can be obtained by further di-

viding group A into two sub-groups:

(a) Those groups which have always adhered closely to

established standards;
(b) Those groups which did not and which still do not adhere
closely to standards.
Figure 3-9 shows that even new improved methods
after
have been found, an important factor for the new methods' ef-
fectiveness is adherence by workers. Furthermore, it is also
 Reliable Method. CEDAC 53

(a) (b)
Groups which focused their Groups which succeeded in
efforts only on adhering to finding new operation
established standards (production) methods
50

40
O

© 20

10

0-10 10-40 40-70 70-90 90 0-10 10-40 40-70 70-90 90

Reduction in Quality Defects (%)

Figure 3-8 Finding New Methods and Adhering to Established

Standards

60-

50

o
54 MANAGERIAL ENGINEERING

shown in Figure 3-10 that adherence increases the possibility

of success in finding new methods.
In conclusion, establishing Category 1 of the New
loharry's Window, i.e., adherence, does not in itself have as
great an effect as finding new nnethods, but the nnethods' effec-
tiveness, as well as the successful discovery of new methods,
will be enhanced by the factor of adherence.
Our next step was to determine necessary conditions for
adhering to standards and finding new methods. The results of
our analysis show that the process of using CEDAC is critical
to both of them.
The eighty-six groups were divided into those which had
succeeded in finding new methods and those which had not.
The distinguishing factor between the groups appeared to be
their approaches to drawing up CEDAC as is shown by Figure
3-10. The successful groups were those which had:
• A high level of adherence to former standards
• Frequent submission of ideas on cards from workers
• More active discussion compared to that of former group
activities
• Frequent revisions of the diagram
Also, is it interesting to note that the successful group
had sufficient leadership and training for the newly estab-
lished operations. That is, after successfully challenging Cate-
gory IV, they switched their attention back to achieving
Category 1.

Figure shows the necessary conditions for adhering

3- 1 1

to standards. The groups with sufficient adherence were those

which had:
• Frequent discussions
• Frequent submission of ideas on cards from workers
• Sufficient clearly-defined standards
• Adequate worker participation in drawing up CEDAC itself
 Reliable Method: CED AC 55

Q>
56 MANAGERIAL ENGINEERING

c
 Reliable Method: CEDAC 57

It should be noted that this information, drawn fronn ac-

tual experience, shows quantitatively how essential it is to
have both the participation of all persons concerned and
mutual trust among all of the group members in order to pro-
mote the adherence to standards and the constant develop-
ment of improved methods.

3.11 CREATING A CEDAC MODEL

Next we developed a CEDAC model which can be used to fore-

cast the results of ongoing CEDAC activities. Figure 3-12
shows the steps by which the model was formulated.
The CEDAC model has two objectives:
• Forecasting the possible outcome of ongoing CEDAC
activities
• Helping us determine which factors should be controlled
in order to obtain the most desirable result under any
given conditions
Only part of the CEDAC model and our research results
are described below. The entire study was detailed in Hyojunka
to August-Sep-
Hinshitsukanri (Standardization and Quality Control;
tember 1978, Japanese Standards Association). This paper was
awarded the prize for best QC paper by the Deming Prize Com-
mittee in 1978. A summary was also carried in Quality Progress,

the journal of the American Society for Quality Control.

A tool, if it is to be useful, must achieve tangible results.

Figure 3-13 shows the percentage of losses due to quality de-

from 1976 to 1979. It decreased by approxi-
fects against sales
mately 50%. These results were not obtained exclusively
through CEDAC, but rather through company-wide QC ac-
tivities of which CEDAC was the backbone.
58 MANAGERIAL ENGINEERING

STEP1
Investigation of eighty-six cases by a questionnaire

STEP 2
Calculation of the correlation coefficient between
each pair of the questionnaire items

STEPS
Distribution of the questionnaire items, using the
coefficients as distances between two items

STEP 4
Division of the items, from their distribution,
into three clusters

STEPS
Distribution of the attributes of each item cluster

STEP 6

—
Formulation of four measures psychological environment,
production environment, effort and participation
from the distribution of the attributes

STEP 7
Calculation of the scores of each of the eighty-six groups
using the four measures, and distribution of the groups
by their scores

STEPS
Division of the eighty-six groups by their distribution
into four patterns.Examination of the relationships among
the four patterns, and the results of their CEDAC activities

Figure 3- 1 2 Formulation of the CEDAC Model

 Reliable Method. CEDAC 59

3.12 GROUPING OF THE OUESTIONNAIRE ITEMS

Twenty-three items which influenced the results of CEDAC ap-
plications were selected. From the correlation coefficients of
each pair of the twenty-three items it was found that these
items fall into three groups:
(1) items concerning the production environment such
as defect frequency and the number of operations producing
defects;
(2) Items related to the psychological environment, such
as concern shown by managers and motivation for solving
problems;
(3) items regarding the use of CEDAC itself, such as revi-

sion of the diagram and submission of ideas on cards.

The degree of influence of each item was calculated in

0.8

0.7

Q en
cn^ 0.6
"5
^ en
,^

o en
- o
^B 0.5

Bc i^>>
0.4
O 13

Q.

0.3

T.
1976 1977 1978 1979

Figure 3-13 Decrease in Quality Defect Rate

60 MANAGERIAL ENGINEERING

Steps 5 and 6. Figure 3- 4 1 shows the results of this calculation

for items related to the production environnnent. Attributes
with a more positive index have a more favorable influence on
the production environment. Items characterized by a rela-
tively large range between positive and negative attributes
show a greater influence on the production environment.
Weobserved that the single most influential attribute is
sufficient adherence to former standard, which again supports
the hypothesis of the New loharry's Window. Second in influ-
ence is management's commitment to emphasizing product
quality rather than efficiency. The second item, defect rates, is
significant, and shows that a lower frequency encourages
more appropriate actions. This means that reducing the defect
rate improves the production environment and facilitates
further defect reduction. The timing of defect detection is also
important. If defects can be spotted without delays and with-
out hampering operations, solving QC problems becomes
much easier. Devices enabling early detection are quite help-
ful.

items pertaining to the psychological environment are

similarly analyzed in Figure 3-15. Influential items include the
group leader's experience, motivation problems for solving
(which supports my contention that we should choose prob-
lems we are eager to solve in our own work), the leader's view
CEDAC, workers' views
of the reliability of of the reliability of
CEDAC, and the concern of managers.
Regarding item 3, it is interesting to note that a nearly
equal ratio of veterans to new workers produces the least de-
sirable effect. The items referring to leaders' and workers'
views of CEDAC's reliability reflects their expectations of
CEDAC's usefulness as a tool for dealing with serious prob-
lems.
Although the influence of Item 9, the range of CEDAC ac-
tivity, is not in itself extremely significant, another it does raise
important point. Suppose the overall probabilityof success for
 Reliable Method. CEDAC

i Poor Good I
1 Relative Emphasis on Quality and
Efficiency
• quality emphasized less
'
than efficiency ^

• quality emphasized as much /

q qq5\
as efficiency
• quality emphasized more 1
0.392)
than efficiency

2 How Often Defects Are Discovered

• daily (-0.136)

• a few times a week (0 172)

• less than a few times a
0-297)
month (

3 When Defects Are Found

• during operation ( 0.254)

• after operation but during (-0.203)

the process
• during the succeeding pro- (-0.206)
cess

4 Adherence to Former Standard

• sufficient ( 0.450)

• nearly sufficient (-0.045)

• insufficient (-0.184)

Production Environment

Figure 3- 1 4 Scores of the Attributes of Production Environment

62 MANAGERIAL ENGINEERING

1 How Often Leader Performs Oper- i Poor Good I

ation
As often as ordinary workers (0.138)
Less often than ordinary workers (-0078)

Seldom (-0.172)

2 Leader's Experience
The most experienced in his group (0.171)

One of the most experienced ( 0.050)

Less experienced than some (-0.305)

3 Ratio of Veterans to New Workers

in Group
Almost all veterans ( 048)

More veterans than new workers ( 089)

Nearly equal (-0 124)

Fewer veterans than new workers ( 0.014)

Almost no veterans (0.031)

4 Participation of worker group in

former QC activities
Very active (0.146)
Active (-0.017)

A little active (-0.054)

Inactive (-0.107)

5 Motivation for solving problems

Very high ( 0.096)

High (0.118)
Low (-0.322)

6 Importance given to the problem

The most important in the plant ( 0. 1 22)

One of the most important (-0.056)

Less important than others (-0 1 30)

Psychological Environment

Figure 3- 5 1 Scores of the Attributes of Psychological Environment

 Reliable Method: CEDAC 63

7 Leader's view of the reliability of

CEDAC
64 MANAGERIAL ENGINEERING

CEDAC activities is about 80%. If many groups in a plant use

CEDAC, then 80% of those groups will succeed, and people can
correctly assess CEDAC's true effectiveness. on the other If,

hand, only one group in a plant tries CEDAC, and happens to

fall in the 20% failure category, the entire plant will become

discouraged about using CEDAC, and QC activities in that

plant will be considerably delayed. This is exactly what we did
observe in some plants. Trying out a new method in several in-
dependent test cases is a preferable strategy.
Next, when the items directly relating to the use of
CEDAC were analyzed in the same manner, two different kinds

of influence were observed: participation and effort. Combina-

tions of the two kinds of influence revealed four basic types of
CEDAC activity. For example, the "little participation/much ef-

fort" type occurs when only a few people, possibly foremen and
engineers rather than workers, devote all of their energy to
solving quality problems.

3.13 PARTICIPATION BY ALL LEADS TO A

BREAKTHROUGH
From the above analysis, it is clear that CEDAC activities are
described by four parameters:
• Production environment
• Psychological environment
• Participation
• Effort
The between the CEDAC activities thus described and
relation
was analyzed in Steps 7 and 8 of Figure 3-12.
their results
Since the objects of this model are human activities
which are themselves the product of innumerable factors, it is
impossible to predict their results as one can when studying
physical phenomena. However, four distinct patterns of
 Reliable Method. CEDAC 65

CEDAC activities did appear. Furthermore, we analyzed how

successful the groups of each pattern were in finding methods
and in achieving adequate levels of adherence. As a result, it

has become clear that the success of each group is closely cor-
related to the pattern to which
it belongs. Table 3-3 gives a list

of the characteristics of each pattern.

Pattern A. Benefiting from better environments, both
participation and effort were high. The groups were successful
in finding new methods and in achieving high levels of adher-
ence. Results were very satisfactory. Almost all cases where
speedy and considerable results were obtained fall exclusively
into this pattern.
Pattern B. Both environments were poor, and not much
extra time was left for problem-solving efforts. However, par-
ticipation was sufficient. Discovering new methods was
difficult for groups fitting into this pattern, but sufficient
adherence was easily obtained. Fair results were derived as a
result of this adherence.
Pattern C. Working in a fairly good environment, group
members expended much effort in trying to solve their prob-
lems. They were often successful in finding new effective
methods, but their adherence was not good, probably as a re-
sult of limited participation. The results were fair or consider-
able, but slow.
Pattern D. Both the production and psychological envi-
ronments were poor. Neither participation nor effort was
sufficient. The overall results were poor.
From all that has been mentioned above we have derived
the following principles concerning CEDAC activities:

( 1 Groups working in a good environment should aim at

Pattern A. If, as is the case in Pattern C, only a handful of
people work hard to find new methods, a longer period of time
willbe needed.
(2) Groups working in a poor environment should aim at
66 MANAGERIAL ENGINEERING

to
 Reliable Method: CED AC 67

Pattern B, which emphasizes participation because these

groups do not have extra time to spend on problem-solving
efforts. Through active participation, these groups can im-
prove environments so that they also can aim at Pattern A in
the future. In other words, Pattern B is a breakthrough for a
positive circle.

3.14 CONCLUSIONS FROM CEDAC APPLICATIONS

The results of the study made it clear that our work contains
some important and unexpected implications for managerial
engineering itself, and not just for the application of CEDAC.
The following are our key findings:
( 1
It is generally accepted that a critical series of steps
must be followed in the solution of quality control problems:

analysis, improvement, standardization and adherence. In the

above it has been proved that adherence is more than a means
to achieve quick and significant results, it is also a springboard
for developing quality improvement methods. Furthermore, it
is a method for ensuring and increasing the effectiveness of

newly discovered methods.

In other words, it is necessary to go full circle in Figure 3-

1 6 from complete adherence to the manufacturing and operat-

ing standards which are already known. This must be repeated
until the percentages of quality defects are reduced to satis-
factory levels.
(2) sometimes felt that management is caught in a
It is

bind between participation, which is a slow process, and

downright elimination of quality defects. The data we have ob-
tained show that these two goals can be fully integrated. The
participation of persons concerned is indispensable in
all

achieving complete adherence as well as finding new

methods.
68 MANAGERIAL ENGINEERING

1 ADHERENCE

4 STANDARDIZATION
\
2 ANALYSIS

V 3 IMPROVEMENT
y
Figure 3-16 The Adherence-Improvement Cycle

It is wrong to eliminate participation on the grounds that

it takes too nnuch time. On the contrary, it is necessary in order
to bring about solutions in a short period of time. This has
been demonstrated in our efforts to solve quality control prob-
lems. As a final remark, it should be added that participation
should always be directed towards a clear-cut, concrete objec-
tive.

An ancient Chinese thinker, Han Fei, wrote:

A king once asked a painter. "V^hat are the most difficult things
to draw?" The painter replied. "Dogs and horses." "Then what are
the easiest?" asked the king. The painter's answer was"hAonsters."
\t difficult to draw ordinary animals such as dogs and horses be-

cause everyone knows them and watches them every day, which
means the painter cannot paint them carelessly. Monsters, on the
other hand, are easy to draw because no one has actually seen

them, and the painter can draw them any way he likes.

Since began my career in the Research and Development

1

Division, have often said, half in jest, that it is more difficult to

1

develop managerial engineering than to develop production

engineering and other R&D. The objects of managerial en-
gineering (quality, productivity, production control, etc.) are
obvious to everyone. Even people in marketing and personnel
divisions have something to say about them. Non-specialists
always find it easier to talk about managerial engineering than
 Reliable Method: CEDAC 69

about such technical subjects as R&D.

Painters of dogs and horses must exercise great skill if
they hope to sell their works; drawing what everyone knows
well is difficult. Fortunately for these painters, however, they
may actually and easily observe their subjects. Ifthey observe
carefully, they note fantastic lines and colors. Having actual
objects to draw is a great source of strength for this kind of
painter. Similarly, detailed and analytical observation is the
first step in managerial engineering.

3.15 QC FOR THE CONSTANT PRODUCTION OF

HIGH QUALITY PRODUCTS

The study of CEDAC sheds light on a critical point, for the im-
plementation of quality control. We must achieve a state
where manufacturing and operating standards are known and
practiced by everyone. QC activities provide the means by
which we may achieve this state and thus secure ever higher
levels of product quality (Table 3-4).
70 MANAGERIAL ENGINEERING

They include functions not only for maintaining standard

them and eliminating defects
conditions, but for restoring
when deviations occur.
Table 3-4 was developed by analyzing and synthesizing
fifteen actual cases where defects were drastically decreased
by achieving complete adherence to existing standards. A vet-
eran foreman once said: "Every item on this table represents
exactly what we have been working hard for. Our daily efforts
are directed towards realizing these four items." It is no won-
der that defects are tremendously decreased when any one
item on this table is realized.
In Table 3-4, item (i) refers to any part of the equipment
that may influence product quality. If product quality is pres-

ently at a satisfactory level, then the state of every part of the

equipment should be maintained as is. This means keeping a
high level of uniformity through thorough maintenance. With
regard to the part of the process which involves human work-
ers, item (ii) aims at achieving a state where effective standard
operations are known and correctly practiced by every worker.

3.16 THOROUGH PRACTICE

One of Sumitomo Electric's plants has been able to com-
pletely implement items (i) and (ii). As a result, perfect pro-
duction quality has been the rule since March, 1976. Perfect
production quality has been defined at Sumitomo as a state
where no quality defect appears in any of the total processes of
a plant. The plant in question achieved this perfect result
through complete equipment maintenance and thorough
training of workers.
The Ouality Control Problem Study Group investigated
how perfect production quality was achieved at this plant. The
possible factors were listed on 400 cards which are sum-
marized in Table 3-5. This book does not attempt to give the
 Reliable Method: CED AC 71

full details of the study, but rather to suggest to the reader that
zero-defect production system was realized by reliable
methods developed within the plant itself.

should be noted here that there is a critical factor for de-

It

veloping reliable methods. This is the power to thoroughly

practice reliable methods. Dr. luran once said: "1 have recom-
mended quality control in the U.S. and in other countries, but
it was only lapan that practiced my idea immediately." Man-
agerial engineering can succeed only when reliable methods
are reinforced through constant practice.

3.17 QUALITY COST

Our approaches to were introduced

perfect production quality
by the plant manager. His emphasis on education and training
was evident: there were times when man-hours spent on off-
the-job training reached almost 7 to 8% of total man-hours.
The concept of quality cost was first expressed in Total Quality
Control by Armand V. Feigenbaum (McGraw-Hill, New York,
third edition 1983). It is a standard used for managerial deci-
sion making and the resolution of quality control problems.
"Quality cost" is defined as the ratio of total costs spent for
quality control to sales, including prevention costs (QC educa-
tion, etc.), appraisal costs (inspection and experiments, etc.)
and failure costs (losses due to defects and claims, etc.). in
terms of this concept, a plant without defects may not be
maximizing its profits. don't think that every plant should ap-
I

proach perfect production quality. Still, the concept of perfect

production quality is helpful. It points out a desirable direc-
tion and leads to a competitive advantage that cannot other-
wise be readily realized. Plant A's quality cost data are shown
in Table 3-6.

The Quality Control Study Group attacked the problem of

quality costs head-on by challenging the widespread tendency
72 MANAGERIAL ENGINEERING

1 Maintenance of work environment. • Use spare time for study meetings

• Keep floors and equipment clean • .Use morning meetings and other opportunities
• Keep everything In order to exchange necessary information
• Train operators to repair equipment by them- • Encourage workers to learn by themselves, and
selves to experience the delight of successful im-
• Keep a low noise level (Slowing down motors provement
may be necessary )
• Put emphasis on the process of workers Im-
• Check and maintain pass lines and detectors provements rather than their results

2. Consistency in fundamentals. • Evaluate workers with multi-dimensional

• Be punctual measures, and group workers with complemen-
• Do physical exercise and running tary abilities and characters together
• Be a reliable person and develop trustworthy
relationships with others 7. Paper tests for workers.
• Practice safety tests, prepared by the group
• Encourage foremen to take initiative
leaders
3. inspection of equipment. • Practice standard tests to encourage workers
• Maintain equipment at peak condition (Do to learn operation standards (Standards are
not overlook even a loose screw |
improved through these tests |

• Organize the Inspection patrol, which consists

of the plant manager, foremen, and leaders of
worker groups 8. Operations standards.
• Establish practice standards
• Promptly follow the advice of the inspection
• Describe key points and know-how clearly so
patrol
that each worker can follow them easily
• Let the plant manager take the initiative In
inspection patrol
• Put cards which visually show key points and
• Make inspection in all areas, e safety, quality,
i
know-how on all equipment
and productivity
9. immediate detection and complete elimi-
4. Elimination of dangerous operations. nation of equipment failures.
• Make safety top priority • Establish preventive measures for equipment
• Warn each other by writing safety memos and failures through the examination of mainte-
safety ratings nance records and with the cooperation of the
• Organize the safety committee and encourage equipment designing staff
it to patrol • Stop machines if there Is any possibility of fail-
• Utilize warning memos and safety campaigns ure (Operators themselves should ask the help
• workers and develop
Utilize safety ideas of fail-
of maintenance workers )

proof and foolproof systems (Standards • Organize an operators' equipment course to

should be easy to adhere to |
teach them about the mechanics of the equip-
5. Quality assurance in each process. ment
• Be responsible for the goods you process
• Let process checkers be responsible for in-pro- 1 Emphasis on systematic preventive
0.
cess quality control maintenance.
• Prevent the recurrence of defects • Emphasize preventive maintenance
• Utilize automatic detectors effectively • Systematize maintenance operations, and then

• Send defective goods back and help those re- improve their efhciency
sponsible to understand the cause of the de- • Use a shock pulse meter to predict bearing

fect deterioration
• Eliminate all defects in a pilot production • Use the card system for Inspection of equip-
• Take economical measures to solve defects ment (The man-hours needed for maintenance
• Let workers experience the benefits of statisti- operation should be managed by this system )

cal methods rather than dwell on the minute

details of the methods 1 Leaders of worker groups active as front-
.

• Impress workers with the importance of quality line managers.

• Assign targets to each group and check the
6. Complete personnel training and informa-
tion exchange. results (targets such as amount of cost reduc-
• Train new workers one-to-one. on the job tion, work efficiency, failure and repair cost,
Teach them how to do It, let them do It. and loss rate, attendance rate, amount of overtime
evaluate their work Train new employees until work, and number of safety improvement
they can respond well to accidents Ideas)

Table 3-5 Realizing a Zero-Defect Production System: A Summary of

 Reliable Method: CEDAC 73

• Let group leaders report on their group results 1 4. Use of interchange training to broaden
at aperformance review meeting in the plant skill development.
(The results should also be reported within the • Establish a flexible manpower allocation sys-
groups )
tem to meet the variable production level
• Through the cooperation of groups, aim at high • Broaden and improve every workers skill
performance for the entire plant • Expand interchange training to include
• Encourage group leaders to participate, start- maintenance workers and inspectors
ing from the initial planning stage • Record every worker's skill using the skill table
• Assign each group leader a task concerning the which describes the operations he can perform
entire plant, and encourage him to follow up and evaluate the level of his ability
on that task (Possible tasks are quality, pro- • Encourage workers to participate in inter-
duction engineering problems, training, im- change training
provement ideas, safety, work environment, • Use interchange training for teaching workers
productivity, equipment, and materials | how to operate the crane and the forklift
• Organize group leaders through the division of
15. Continual, steady improvement.
tasks, and provide a back up staff system
• Improve quality and efhciency Reduce loss,
• Improve the quality of group leaders through
scrap materials and the number of operators
the measures mentioned above, until, for In-
• Improve equipment Use brand-new equip-
stance, they can talk to the accounting staff
ment efhciently Fully utilize the equipments
knowledgeably
designed capacity
• Establish a computer system that can be used
12. Respect for the creativity of the line by workers
people (foremen, group leaders, and work- • Encourage workers to make value analysis
ers.) ideas and change design specihcation if
• Provide as many opportunities as possible for needed
the line to think by themselves
1Utilize an idea system for steady improve-
6.
• Rely on the lines creativity
ment.
• Assign a part of the staffs task to the line
• Let group leaders quantitatively estimate the
• Provide an environment in which every person
results of the workers Ideas Teach them calcu-
can develop his ability and no one is left be-
hind
lation methods for estimating the results
• Encourage the staff to back up the line
• Utilize the idea bank and idea note
• Encourage workers to make direct contact with 1 7. Perfect quality production achieved
the plant manager in case of emergency through the enthusiasm of all, and maintained
• Utilizeall theopportunities. such ascompany- in a relaxed atmosphere.
wide study groups, to encourage improvement • Work to achieve perfect quality production hrst
• Utilize meetings to get new ideas from the line inthe process with the most difficult problems
(Do not criticize poor results, but with the line, (Success here will increase the enthusiasm of
analyze why they could not achieve good re- allpersons |

sults )
• Maintain perfect quality production in a re-
• Put emphasis on wise investments for cost laxed atmosphere
reduction, rather than on small savings • Try to solve the biggest problems hrst (Success
• Aim for high performance and high motivation here leaves much time for further problem sol-
through rationalizations, which Is carried out ving I

with respect to all workers • Dehne the measures for effectiveness of

maintenance operation
13. Day-to-day management with the partici- 18. Definition of perfect quality production
pation of all persons concerned. and recognition, by all, of the achieved results.
• Encourage all (including workers) to partici- • Define perfect quality production, including
pate in management and control the problems which bring big losses, and the
• Estimate every day s performance through day- problems which Involve everyone
to-day management • Publicly share successful achievements (for
• Enthusiastically usetheldeasof youngerwork- example, by giving special badges and marks
ers on helmets, taking commemorative photo-
• Emphasize creative use of spare time rather graphs, and planting commemorative trees)
than the rigid efhclency-hrst principle • Utilize the award system for improvement
• Organize work divisions and group divisions ideas, safety Ideas, special tasks, training, etc
effectively • Be patient while working toward perfect quality
production

the QC Problems Study Group's Ideas

74 MANAGERIAL ENGINEERING

^"^••^^^ Period
Cost Items^****^.,^,^
 Reliable Method. CEDAC 75

which produces large electric systems, a product failure might

result in enormous claims from the user. In such cases it is im-
portant to calculate potential failure costs from claims and ac-
cordingly set desirable levels for prevention and appraisal
costs. We
should take into account both actual failure costs
which have been incurred, and possible ones, no matter how
improbable they may appear. Prevention and appraisal costs
should be budgeted on the basis of this calculation. Thus the
concept of quality cost leads to strategic quality control.
The second space shuttle was launched into space in
November, 1981. The mission was deemed a success. have I

heard it said that its cost per pound was equal to that of
platinum. Considering the actual manufacturing costs, pre-
vention and appraisal costs must have been enormous, which
reflects a huge amount of potential failure cost. It is significant
that this potential failure cost had to include many intangi-
bles, such as a potential loss of national prestige, in the event
of a disaster.

3.18 A PROFESSOR FROM DENMARK

A professor of mathematics interested CEDAC visited us in
in
the fall name was Dr.
of 1980. His Dahlgaard. When explained
1

the New Window to him, he exclaimed: "Is it possible

loharry's
for us not to practice what we know? do not understand the
I

known-unpracticed category at all." He spent two days on this

question. At the beginning, we were also puzzled by the
mathematician's viewpoint on CEDAC. However, at the end of
his study, he grasped the essentials of CEDAC. He wrote to us
that he would give a lecture on CEDAC in Copenhagen in Jan-
uary, 1982.
While Dr. Dahlgaard was staying with us, one of our
CEDAC groups achieved an astonishing result which is shown
in Figure 3-17. Even in an automated plant, workers can make
76 MANAGERIAL ENGINEERING
Reliable Method. CEDAC 77
78 MANAGERIAL ENGINEERING

a significant contribution to problenn solving. A CEDAC chart

had been drawn to decrease the frequency of incorrectly rolled
was appearing between stands 21 and 22 in a
steel wire that
rolling operation. "Mis-rolls" refers to defects in this process
caused by nnachine troubles and operations errors. This ex-
perience clearly illustrated the process from fact-finding to
problem recognition and solution. As a result of this activity,
the productivity index of man-hours needed to produce a
given amount of steel wire increased from 100 in October 1977
to 164 in December 1980.

3.19 NECESSARY CONDITIONS FOR RELIABLE

METHODS
Figure 3-18 is taken, with minor revisions, from How to Get Crea-
tive Ideas (Hassoho, 1977), in which Professor |iro Kawakita in-

troduces his K) (QC circle) method for the first time. He pro-
poses that in the analysis of scientific methodology, three
kinds of sciences should be taken into account: field science,
speculative science, and experimental science. 1 reviewed our
group's activity in terms of his W-shaped framework.
First of all, the problem was defined at Point A as the
proposals shown in Table 3- 1 . There compared the ideal con-
1

ditions to be realized with the present conditions which were

to be improved. 1 then proposed our own original method to
achieve the former.
Next, the group members experimented with various
changes in their work to achieve ideal conditions. This was the
exploration process shown asked them, quite
in Figure 3-18. 1

literally, young plant. Choosing a plant to grow

to "grow a '

within one's own work was an essential factor for maintaining

the members' zeal; it is important to keep in mind that mem-
 Reliable Method. CEDAC 79

bers' energies should always be focused on solving their own

critical problems.

The nnembers brought their plants to the group meetings

and exchanged ideas. This was observation B-C at the experi-
ence level. In this process, CEDAC was created by modifying
cause-and-effect diagrams and control charts.
As CEDAC spread throughout the company, many re-
markable results appeared (Table 3-2). The study group was
excited and decided to investigate their causes. The
hypothesis of New loharry's Window was formulated; this was
point D.
In order to explain the hypothesis, eighty-six cases of
CEDAC applications were collected and analyzed (see Steps E,
F and G). In the observation process from F to G, the computer

The Work of Development

Speculative Science

THE THOUGHT LEVEL

THE EXPERIENCE LEVEL

Field Science Experimental Science

Figure 3-18 Scientific Methodology

80 MANAGERIAL ENGINEERING

and statistical methods were used to deal with many factors

and an enormous amount of data.
At point H, the hypothesis was tested and valuable QC
know-how was obtained. We did not forcibly work with this W-
shaped framework in our minds. However, it does perfectly de-
scribe the process of our study. Knowledge and know-how
obtained through these steps can be referred to as "reliable
methods." The study group developed "theories out of prac-
tice" by adhering to these steps of scientific methodology.
It should be emphasized here that the quality of any

hypothesis depends both on speculative science and field sci-

ence. We must be simultaneously practical and theoretical in
our outlook. Rather than trade off these two attitudes, we must
integrate both This is the reason that we must study theory
and improve the quality of our observations at the same time.
If both our knowledge and observation are poor, how will it be

possible for us to develop sound hypotheses?

Statistical methods are used to test these hypotheses. We
can learn how to do this in school or through textbooks, but we
can only learn how to formulate sound hypotheses by our-
selves. Poor hypotheses will not lead to solid results no matter
how good the tests are.
To obtain good results in a plant environment, the inte-
gration of everyone's ideas is necessary. This was facilitated by
the study group approach. This kind of integration was also
realized among various study groups Figure 8-
(
1 ), among vari-
ous field sciences, as well as between field science and
speculative science.
chapter 4

A RELIABLE METHOD (2):

STOCKLESS PRODUCTION

4. 1 PRODUCTION CONTROL — BASICS FOR THE

MANUFACTURER
In industrial plants, production is carried out according to pro-
duction plans. Establishing these plans is a basic daily taskfor
the manufacturer. The goal is for materials to flow smoothly
through processes until the final product is produced with de-
signed quality.
In practice, production control is aimed at several objec-
tives:
• To minimize the total man-hours needed for processing,
or, in other words, to maximize output per man-hour
• To minimize total delivery delay
(Most plants try to realize these first two objectives
simultaneously)
• To minimize total production costs
• To minimize total work-in-process time
Does your plant prepare production plans with any of the
above objectives in mind? Aiming at all of them at once is
theoretically impossible, and it is the workers who are caught
in the vise. Plants must establish standards in order to balance
out these conflicting objectives, or to achieve one of these
goals at the sacrifice of others.

81
82 MANAGERIAL ENGINEERING

The techniques of Operations Research (OR) enable

plants to achieve any one of the four above-mentioned goals.
Unfortunately, these techniques are impractical, because the
size of the problems they can handle is too limited in scope
(several jobs and several machines at most) compared to the
actual problems faced even in a small plant which might com-
prise more than a thousand jobs and machines. However, un-
derstanding these relatively simple cases is indispensable for
grasping the essence of the problem, and herein lies the true
value of operations research.
Problem size and other factors make it difficult to apply
the techniques of operations research in actual situations.

Another factor is the need to know how the optimum solution

will be affected by changes in the environment, something
that is not easily determined by simple OR models. Further-
more, there are many cases where it is necessary to achieve
several objectives simultaneously.

4.2 DEVELOPING THEORIES OUT OF PRACTICE

On October 25, 1944, the carriers of the U.S. Navy Task Force 5

were suddenly attacked by kamikaze fighters near the Philip-

pines. The Navy had been continuously harassed by them for
the last ten months of the war. Recognizing the seriousness of
the problem, a conference was held at San Francisco from
November 24-26, 1944. Several defense organizations partici-
pated conference and analyzed together all the availa-
in this

ble materials and possible countermeasures. A group of oper-

ations research specialists was also involved.
By analyzing how the fighters attacked and how American
ships responded, the operations research group was able to
find a solution for decreasing the rate of successful attacks.
This solution was recommended to the Navy. Some ships
adopted it, but others did not. Thus the solution was uninten-
tionally tested during actual battles. Those ships that had
 Reliable Method. Stockless Production 83

adopted these tactics decreased theby 50%. hit rate

hesitate to cite this serious case, but it demonstrates the

1

effectiveness of operations research. The nnethodology for de-

veloping managerial engineering techniques, i.e., "developing
theories out of practice," can be observed in this case. Or,

rather, the methodology is operations research itself. Whether

the objective be quality or productivity, this inductive method
for developing managerial engineering can in fact be called
OR.
As we said in Section 4. OR models are often too limited
1 ,

and simplistic to solve practical problems. But my under-

standing is that OR constitutes an attempt to simplify and
model a complex phenomenon so as to clarify its essential
structure. By constructing a clear (and often mathematical)
model, we are able to deepen our understanding of the
phenomenon. The model's role is not to describe every detail,
but rather to provide a better basis for decision-making.
As an example of a mathematical model, consider the fol-
lowing which is used as a decision rule to establish the
maximum electricity allowance (an explanation of the vari-
ables is omitted):
Wi
>1.1
Wo

(Bi+O. 9(B2-Bi)+0. 7(^^—B2)]d

W= Wi'- ^, -^ i3

How would be possible to describe such a complex rule

it

without recourse to mathematics? In the business world

today, we
becoming accustomed to using similar
are
mathematical models rather than redundant verbal expres-
sions.
Through the use of these principles, we decided to start
with an understanding of the structure and functions of the
84 MANAGERIAL ENGINEERING

complex phenomenon (industrial production) to

develop ef-
fective production control systems. We learned this approach
of starting with a basic understanding of the system from
some European firms visited in the summer of 1977. The writ-
ings of a famous Japanese IE researcher, Mr. Shigeo Shingo.
were also quite suggestive.

4.3 MOVING TOWARDS STOCKLESS PRODUCTION

At the time we were beginning our efforts, Toyota's kanban
system was rapidly spreading throughout Japan. My intention
was to transform kanban into an engineering approach that
could be practiced by anyone in any plant. With this end in
mind, the Production Planning and Control Study Group was
formed in October 1977. We also started an IE Applications
Study Group with volunteers from other companies. The prin-
ciples of stockless production were developed from the prac-
tice and study of these two groups. Strictly speaking, it should
be called "less-stock" production, but this does not sound as
good as the term "stockless."
Stockless production is a method for organizing opera-
tions that aims for maximum efficiency. This means producing
necessary products in necessary quantities exactly at the time
they are needed, it emphasizes the critical role of stock, or
work-in-process, and attempts to develop planning and con-
trol systems for smooth material flow. It also helps realize this

goal by suggesting improvements in production systems.

The greatest obstacle to obtaining ideal production con-
ditions is learning how to approach the vastly complex prob-
lem of inefficiency. The stockless concept provides a key to this
approach.
We shall use the term process to refer to materials flow; the
transformation of materials by man or machine is termed oper-
 Reliable Melhod. Slockless Production 85

ation. If management attempts to achieve both rapid ma-

plant
from the perspective of process and minimize idle
terials flow
work-time from the point of view of operations, a conflict will
arise.
The only way to resolve this conflict is to constantly im-
plement improvements in the production system. The follow-
ing data show that these two different objectives of process
and operation can be made compatible by improvements. Fig-
ure 4- 1 illustrates the percentages of stock reduction that have
been obtained in stockless production applications in differ-
ent types of manufacturing companies. Work-in-process was
reduced by 48% on the average. As a result, manufacturing
lead time was also reduced as shown in Figure 4-2. Productiv-
ity improvements simultaneously achieved in these cases are

also shown in Figure 4-3.

The point is to make the contradictory objectives of pro-
cess and operation compatible. The key parameter of this
problem is the allowable stock between two consecutive steps
in the production process. The term "stockless" was derived
from this idea.
"Allowable" refers to the maximum stock at any point in

the process. In general, of course, this is only the upper limit;

constant effortis exerted to keep stock levels as low as possi-

ble. By controlling stock quantity, the characteristics of a

stockless (just-in-time) production system are determined.
Management sets this allowable quantity — zero, or a three-
day supply, or whatever — taking into account how quickly im-
provements can be made.
Setting the allowable quantity too high does not make
sense since the very existence of excess stock levels tends to
disguise production problems. It is only when the plant is con-
strained to work with minimal stock levels that inefficient as-
pects of the production process leap to the surface, for all to
see —
and correct. If, on the other hand, the level of allowable
86 MANAGERIAL ENGINEERING

u
(O

E
3
 Reliable Method: Stockless Production 87

quantities is set too low, necessary improvements cannot

catch up with the overwhelming number of problems that crop
up. This inevitably leads to workers secretly exceeding the al-
lowance. In such a situation, improvements are actually de-
layed, contrary to management's intention.

4.4 PRODUCTION STANDARDS AS MANAGEMENT

POLICY

Management should express its policy in terms of standards

such as standard flow time and standard stock quantity. These
standards need not be very sophisticated, but they must be ra-

tional. If production standards are to be respected by workers

they must be practical and understandable. This is one of the
most important lessons we learned from our experiences.
Although a detailed discussion of stockless production is
not the objective of this book,some explanation of standard
flow time and standard stock quantity is in order.

Standards for Planning and Control

(1) "^ hat are standards?
Standards are used to control material flow between consecu-
tive steps in the production process, and thus to control the
total flow in a plant. Starting from the production deadlines,
material flow between any two consecutive steps is controlled
from downstream to upstream, smoothing the total flow.
The basic task of production control is to determine what,
when and how much to make. To do this, material flow must be
controlled through either of twodimensions: timeor quantity.
A standard established in terms of time is called "stan-
dard flow time." It is the time required for material to go
through two consecutive steps Figure 4-4(a) In most cases it
| 1.

is determined for each machine, taking into account lot

88 MANAGERIAL ENGINEERING

(a) Elements of Standard Flow Time

Previous
process

t
E

One
process
(0
•D
C
B

i
Next
process

(b) Example of Standard Flow Time

Process Standard Flow Time

< 2000 kg^ > 2000 kg
A drawing 1 day 2 days

B furnace 2 days

< 1000 kg < 2000 kg > 2000 kg

C drawing 2 days
1 day 3 days

'Standard flow time depends on the lot size

Figure 4-4 Standard Flow Time

 Reliable Method. Stockless Produclion 89

sizes. Quantity refers to production speed and is expressed in

terms of production quantity nneasured in units per day. It is

termed "standard production quantity."

(2) The role of standards
Standards serve for control as well as planning:
• By comparing changing production conditions with stan-
dards, each process can be understood in terms of what
should or should not be done at each moment,
• As workers try to adhere to standards, problems in the pro-
duction process are detected. They are swiftly recognized
and solved by all of the persons concerned.
Therefore, standards serve as the basis for control and im-
provement. In plants where only a limited variety of products
are manufactured, material flow can be grasped in terms of vis-
ible stock quantity instead of flow time. The standard stock
quantity is determined as follows:

Standard Stock Quantity =

Standard Production Quantity x Standard Flow Time

The two systems using standard flow time and standard

stock quantity are compared in Table 4- 1 . In fact, many plants
do not measure materials in terms of stock quantities. Plants
producing a great variety of products fall into this category.
This type of plant cannot introduce Toyota's kanban system
which uses stock quantity to establish standards. But if flow
time is used for standards instead of quantity, the same kind of
control is possible. Stockless production is therefore possible
in any kind of plant.
Standard flow time or stock quantities determine the
characteristics of a production system. Using these standards
as control parameters, one can develop an actual production
system. Detailed rules are secondary problems; one can set
90 MANAGERIAL ENGINEERING

Standard work-in-process
System Standard flow time system
quantity system

Applicable to Make-to-order Make-to-stock

Standard Standard flow time is deter- Standard work-in-process

mined from manufacturing quantity is determined by
man-hours and allowable the capacity of a process
delay. and the demand fluctuation
Divisions of standard flow of the subsequent process
time should be within the Standard work-in-process
span of control of foremen quantity is often used in-
and workers stead of standard flow time,
Starting date of each pro- because work-in-process
cess is determined by the quantity can be visually
delivery date and standard controlled
flow time

Control method Production schedule for Production order is placed

each order is planned by the to keep work-in-process
use of standard flow time quantity at a hxed level

Symbols
[A Flow of Product A,Oct. 3
Product H2
(Control card]
1^^^^^^^ ^^^^, 10 pieces
L/ material P2 B Oct. 5

li ..
\
_ Flow of
information

Control card prepared for

each order accompanies the
material

table of standard flow time

\s.,^Process
 Reliable Method. Stockless Production 91

them at will, it may not be easy, but this is not critical as long
as rational standards reflecting management policy are set.

4.5 THE CHARACTERISTIC APPROACH

It is important for those who practice managerial engineering

to have a clear idea of what they are doing. If a method devel-
ops out of a clear understanding of basic facts, it will lead to
significant productivity improvement. The lack of understand-
ing isoften a cause of failure when a method is blindly applied.
As the work of the Production Planning and Control Study
Group proceeded, stockless production was applied in many
plants of Sumitomo Electric and its affiliated companies. Sev-
eral other firms reported that the production stock quantities
were significantly reduced by following stockless production
procedures.
In luly 1978, just a year after the start of the Production
Planning and Control Study Group, the IE Applications Study
Group was formed at the Kansai Institute of Management and
Information Sciences with fifteen volunteers from other firms
and universities. I was appointed chief of this group. This was
an excellent opportunity to gather information on stockless
applications in other companies. The group was run on the
idea that the basic principles of stockless production can be
applied in any type of plant, in spite of the fact that actual
methods will vary from one application to another.

4.6 EBB TIDE AT THE SEASHORE

Figure 4-5 shows the relationship between stock levels and

various production problems. When the stock level is high, it

covers production problems just as a high tide covers reefs.

92 MANAGERIAL ENGINEERING

[With large work-in-process quantity]

In-process quantity level

^W'
 Reliable Method. Stockless Produclion 93

When production stock levels decrease, production problems

crop up like reefs appearing one after another at low tide. If
one has sufficient will to make improvements, stock level re-
ductions can become a powerful technique for detecting and
solving problems.
There is a problem wherever production stock accumu-

lates. Reasons for this accumulation are summarized in Table

4-2. The seven reasons given in the table interact with each
other, and thus increase production stock in many cases.
There are other cases where no clear reasons are de-
tected. For example, if a shop operates one shift a day, while
the downstream shop operates on two shifts, work-in-process
for one shift will suffice. There are many cases however, where
three or four days worth of stock is accumulated. In the testing
or packaging process, it is often observed that a handful of
workers are swamped with work-in-process. In such cases.

Reasons for WIP Accumulation

94 MANAGERIAL ENGINEERING

there may be no valid reasons for piled up stocks. But even in

these cases, there are real problems that cause work-in-pro-
cess accumulation. Stockless production does not hide these
problems under high levels of work-in-process, it tackles them
by using low stock levels to bring them to the surface.

4.7 BUILDING A BASIC MODEL

I introduced my ideas on OR models at the beginning of this

chapter. Because of these ideas, we wanted a basic model that
would enable us to understand complex phenomena of pro-
duction control and to develop a control system based on this
knowledge. It is desirable for such a model to be used for ex-
periments that will produce visible facts rather than abstract
knowledge because only facts obtained through experience
can generate new ideas when we are confronted with new
phenomena. The thirteen principles enumerated in Table 4-3
were developed as a model to serve this purpose.
Principles to 6 were obtained from simulations created
1

by the lapan Management Association and first described by

Mr. Shigeo Shingo in one of his books.
Let me briefly introduce the simulation.

Simulation Rules
• Figure 4-6 is the form used in the simulation — the
arithmetic tool should be used as specified for each
problem
• Eight people perform this simulation
• People are referred to as workers, arithmetic tools as equip-
ment, and calculation as operation
• Items 1-lV are referred to as products — one sheet of paper
contains four products
• Eight sheets are used in the simulation — therefore, the
total number of products to be made is thirty-two
• By inserting a different number in the parentheses of each
 Reliable Method. Stockless Production 95

Principle I

There are two dimensions of production, which cross each other:

'

"process" and "operation

Principle 2
Delay occurs as the result of division of labor, and transportation occurs
when a process is divided into subprocesses.
Principle 3
There are two kinds of delays; "lot delay" and "process delay " (Lot delay
refers to the delay of one itenn while other items of the same lot are
processed. Process delay occurs when a whole lot waits between
processes )

Principle 4
The greater part of manufacturing lead time consists of delays
Principle 3
Process delayis considerably reduced by balancing the line and

diminishing bottlenecks
Principle 6
Lot delay is reduced by the division of lots

Principle 7
Limiting in-process material to smaller quantity is an effective means
for shortening manufacturing lead-time.

Principle 8
Balancing the line is effective in shortening of manufacturing lead time
and, at the same time, in raising efficiency

Principle 9
Whether a line is well-balanced or not, efficiency plunges to a much
lower level if in-process material is totally eliminated (i e no delays) ,

Thus, the optimal number of in-process units should be determined

deliberately in order to balance efficiency and lead time

Principle 10
Capacity should be immediately adjusted to the fluctuation of load
in order not to increase lead-time (or process delay).

Principle 1
Waiting time before production begins should be adjusted by
controlling backlog

Principle 12
Shortening the planning cycle leads to the reduction of waiting time
Principle 13
Maintenance of stock of semi-finished products is an effective means to
shorten manufacturing lead time of made-to-order products However,
the stock should be kept at a minimum

Table 4-3 Principles for Production Control

96 MANAGERIAL ENGINEERING

ITEM I ITEM m
f g
3+4+ 2-^9 ^ 7+l2^25*( )

0x 6-1-8 ^ I X 2 -10
)-!- 14 5 <[)-'• 2 12 -176 ^ 0-K2)+42 5-1-182-504
@x6- 15 ^ A @^ 2 X

l@-f<T>(g) f 85
1

5)' @-^@*(D-240^^(D •

(i)lMl®x2 M®
(X)|m ©•<-4" © M ©-^6
S l(6)x®-25 ^S (7)x(g)^ 120"

ITEM n
 Reliable Method. Stockless Production 97

paper, we can change the arithnnetic problem

• Processes through © should be carried out in order.
• Numbers (D through in each problem indicate that
the results of previous calculation should be used accord-
ingly.

Simulation Outlines
The following five simulations are conducted in order.

Simulation 1 : Before Division of Labor

Operators [> A
98 MANAGERIAL ENGINEERING

Then he gives it to Operator D, who works on each Problenn

2. Operator A starts calculating each Problem on the sec- 1

ond sheet. This procedure simulates the lot production

with the lot size of four.

Simulation 3. WitA Consideration of Workers' Skill and ^ork Flow (Lot

Production]

c
 Reliable Method. Stockless Production 99

completing his work, the operator hands the paper to the

next person.

Simulation 5: After Line Balancing

c
100 MANAGERIAL ENGINEERING

SIMULATIONS

1 Before Division of
Labor
Each of 8 workers
conducts all processes
foran entire lot. A lot
consists of 4 items.

2 After Division of
Labor
Each worker conducts
one of 8 processes for
all 8 lots. All 4 itenns of
4 effect of
each lot are trans-
diminishing
ported together.
bottlenecks

3 With Consideration
of Workers' Skill
Division of labor is

arranged based on
workers' skill.

4 After Division of
Lots
Each item of the lot is

transported separately

O)

5 After Line Balancing

Work load for each
process is averaged by
tranferring workers
from processes with
idle time to bottlenecks
Reliable Method. Stockless Production 101
102 MANAGERIAL ENGINEERING

The remaining principles were obtained from the study of

the balance between load and capacity using input/output
charts as well as the analysis of manufacturing lead time.
The thirteen principles in Table 4-3 obtained by the study
group produced our model for production control. Although
the model may be incomplete, it enables us to see the essen-
tialaspectsof any production control phenomenon. There will
always be a big difference in results, depending on whether we
have a basic model or not for solving production problems.
Although am not happy to remind you of war battles,
I

please think about the results which the OR model obtained

against kamikaze fighters. Those ships that did not employ the
model suffered twice as many hits as those that did. This fact
should be a powerful lesson to all who are concerned with in-
creasing plant productivity and product quality.

4.8 APPROACHES TO STOCKLESS PRODUCTION

The Production Planning and Control Study Group and the IE
Application Study Group developed the basic approach to
stockless production, as follows.
(
1
Stockless production must be based upon the princi-
ples of production control. Although characteristics of prod-
ucts and production from company to company, the
differ
basic principles of production and work-in-process remain the
same. First of all, it is necessary to understand the thirteen
principles (Table 4-3) in order to develop a stockless produc-
tion system suited to each company's characteristics.
(2) Delay must be avoided within the total framework of
the production system. It is essential to think in terms of a con-
trol system for the entire production system rather than for

any single shop or process. This system will help detect the
problems which must be solved. A standard delay in flow time
or work-in-process quantity used in the control system is a de-
cision variable that determines the required swiftness of im-
 Reliable Method. Stockless Production 103

provements. Such a control system clearly shows the priority

of production problems to everyone concerned. When this
goal is established for everyone, timely and swift improve-
ments in productivity are achieved.
Improvements in the QC and IE activities, the produc-
(3)
tion system,and the production control system must be pur-
sued simultaneously. To obtain stockless production, im-
provements to remove obstructing problems as well as effec-
tive production and production control systems for the
smooth flow of materials are essential. Stockless production is
impossible without these three factors (see Figure 4-9).

(4) The target level of stockless production must be

determined by management of the production division. The
process objective is given first priority. Nonetheless, the oper-
ation objective must also be kept in mind, and various im-
provements, such as set-up time reduction, equipment
maintenance, and quality control must be made accordingly.
Therefore, the target level and approach speed of stockless
production must be determined with due consideration for
make the necessary improvements. Man-
the plant's ability to
agement of the production division must show firm determi-
nation and a clear conception of its ability to achieve these
goals.
Production control through total participation must
(5)
be pursued. Stockless production cannot be achieved by the
planning and control division alone. All the workers must per-
ceive the system of materials flow and understand what they
are expected to do. In other words, everyone controls and
monitors his own work in stockless production. It is important
for all theworkers to share necessary information and decide
what they should do at each step. Control systems that pro-
vide visual information are very effective for this purpose. By
production control through total participation, stockless pro-
duction aims at avoiding old-fashioned systems with shortage
lists and expeditors.
104 MANAGERIAL ENGINEERING
Reliable Method: Stockless Production 105
106 MANAGERIAL ENGINEERING

4.9 HOW STOCKLESS PRODUCTION WAS

DEVELOPED
There was a basic task to be realized by the manufacturing
firm: achieving smooth production flow. Improvements in this
area had long been desired, yet it was still a new frontier for IE
and OR in our company. Some of the managers in charge of
production control were involved in this task and longed for
even the slightest payoff from their work. They had valuable ex-
pertise gleaned from their vast experience.
Existing IE and OR methods
were quite limited in their
potential practical applications. However, they did provide us
with a basic model for production control, and stimulated our
thinking. Although these factors were nothing new, we did
anticipate integrating them in a new way. The Production Plan-
ning and Control Study Group and the IE Application Study
Group provided the opportunity for this new integration.

First order law: Law of duel or strategy for the weak

m^ - m ^ e X {n^ - n]

m^ enemy's initial workforce

m ; enenny's rennaining workforce
rig : our initial workforce
n : our remaining workforce
e : exchange rate (or weapon efficiency)

Second order law law of concentration effect or

strategy for the strong

Table 4-4 Lanchester's Law

 Reliable Method. Stockless Production 107

New combinations of existing knowledge are an impor-

tant means for reaching new horizons in engineering. For in-

stance, a survey of seventeen newly developed American

weapon systems was conducted, which reported that 95% of
the basic techniques utilized in these systems had been de-
veloped forty to fifty years ago. Furthermore, four of the six
most innovative systems were nothing but combinations of
existing techniques.
A great desire to achieve improvements is indispensable.
We realized that it was essential for everyone concerned to do
his best in order to reach the highest levels of creative problem
solving.

4. 1 SECOND THOUGHTS ON OR
It is widely believed that OR
more effective for prevention
is

than for actual treatment. The law shown in Table 4-4 was for-
mulated by Philip Morse and George Kimball in their famous
book Methods of Operations Research (MIT Press & Wiley, New I.

York, 1951). The first order law is called the "law of duel" or,
from another viewpoint, "strategy for the weak." Here, Hq sricl n
represent our workforce before and after a battle. Therefore,
riQ -n shows our losses in the battle. Similarly rriQ, m and
yyiQ-m are the enemy's initial workforce, remaining workforce
and losses.
If this model is applied to our work, Wq -w is the remain-
ing workload to be done, and riQ-n shows consumed re-
sources to be consumed, such as manpower, costs and mate-
rial. Let e stand for the exchange rate which compares the ef-

fectiveness of weapons or workforce between the two parts. If

e = \. the workforce difference remains unchanged after
initial

the battle.
would like to talk about e for a moment. In our study
1

groups we tried to improve its value. The "reliable methods"

and "practice by all" which we discussed at the beginning of
108 MANAGERIAL ENGINEERING

this book are means for raising the value of the exchange rate,
e. This exchange rate may be used to relate past and present, or
us and a rival company.

The characteristics necessary for managerial engineering

to become a reliable tool are-. ( 1
) It must be applicable by any-
one, anywhere, at anytime; and (2) itmust be so suitable forus
that we are able to use it with total confidence. The latter is not
meant compromise with our present situation. Sim-
to imply
ply, the engineering methods must improve our creativity,

coincide with our management culture, and aim at long-term

improvements.
As have stated again and again, both
1 and (2) above ( 1
)

suggest that the creative use of this original knowledge is in-

dispensable to the successful application of managerial en-
gineering. Copying from other firms always creates weakness.
This point is also important for integration. Successful inte-
gration with the ideas of others must be based on one's own
strengths.
Regarding "practice by all, it was not the insufficient level
"

of our own knowledge that obstructed our effective applica-

tion of managerial engineering. It was our practice that was in-
sufficient. A strong sports team must practice abundantly in
addition to possessing great skill. That which we are unable to
obtain quickly may represent a true competitive advantage.
Managerial engineering obtained through abundant practice
can become a source of tremendous strength.
When five units of a red army and four units of a white
army each other with an exchange rate of one (in, say. a
fight
medieval war), the white army is completely destroyed, and
one unit of the red army survives. In a modern battle, the
number of remaining units' equals 5''-4^ Thus, the red army
will be left with three units rather than one after the battle. ac- 1

tually experienced the principle behind this kind of exchange

in the guise of workload (enemy workforce) versus design

manpower (our workforce) while building a new plant. A plant

 Reliable Method. Stockless Production 109

had to be constructed by my division on a new site, and was in I

charge of the project. It was the first time we had attempted

such a project, and we were very inexperienced. We decided
that since we were using inexperienced people, we would use a
larger workforce —
more manpower —
than was actually re-
quired.
At that time, the division had three other plants, and
available plant engineers for the project were drawn from all

three of them. The project's workload fluctuated, as it pro-

ceeded, from five to three to six. As a result of our policy, we
were able to bring the plant to successful completion in spite
of our lack of experience. Needless to say, the plant managers
alone were not the only reason for our success; many people
inside and outside of the division gave their entire support.
In spite of the fact that the project was a success, would I

have been able to do it with less manpower if had known Lan- I

chester's law — the law governing the exchange between the

red and white armies. should always have employed one
I

more unit of manpower than was required.

The reader may doubt how it is possible to quantify the
workload as just described. reply that it is the manager's (not
I

the computer's) responsibility to assimilate and control such

unquantifiable factors.
If you make it a rule to use four units of manpower for five
units of workload, you should reconsider your policy. Unless
your exchange rate is truly extraordinary, your "armies" risk de-
feat on every front. The Prussian Army's basic strategy, for in-
stance, was to concentrate all the workforce as swiftly as possi-
ble in order to destroy the best forces of the enemy. On ac-
count of this strategy, railroads were more important to them
than fortresses.
I recommend that the reader use OR models informally in
his daily work. It is always easy to find someone who has a
good knowledge of OR and who can provide many constructive
suggestions. There is no need for using formal OR terminol-
110 MANAGERIAL ENGINEERING

ogy. Fruitful thinking is more important. We can not achieve

anything until a problem is first thought out.
The leader must be able to spot scenarios for success by
integrating what he needs to maximize ^(^0 -n). That is, he
must find and show his subordinates a strategy for winning, or,
at least, not losing.
The higher the leader's position, the more important this
task becomes. Nothing is more irresponsible than for the
leader to put his subordinates on the front without any pros-
pect of victory. The leader should develop his scenario for suc-
cess and take the lead by showing it to his subordinates. Need-
less to say, mere hard work is not enough for him.
introduced my idea on motivation in Chapter 2, when
1

discussing the hydrogen accident prevention project. There

are numerous books on motivation now, but believe the true 1

means of motivation lie in the leader's communicating a

scenario for success which can be accepted and believed by all
his subordinates. All other means of motivation do little more
than offset negative factors.
Furthermore, have learned that the most effective means
I

for everyone to accept and believe an idea is understanding

based upon scientific discovery.
In managers do not use OR very effectively, if they
general,
just make the effort to implement its basic principles, they will
be rewarded by increased productivity. Both rational thought
and unexpected inspiration are necessary conditions for crea-
tive work. When one pursues rational thought by building an
OR model, he may very well be led down unexpected paths.
Such unexpected paths lead to creativity.
chapter 5

A RELIABLE METHOD (3):

IE IMPROVEMENTS

5.1 ROBOTS AND IE

New techriologies such as robotics, automation and the in-

creased electronic sophistication of machinery are changing

the world of industry, industrial engineers especially cannot
afford to neglect progress in this field.
The other day watched a publicity film by a robot man-
1

ufacturer. was astounded to observe several times in the film

1

that his robots performed useless motions. thought they

1

were funny and felt a sort of affection for these almost-human

robots. 1 learned, at the same time, that the efficiency of a
robot depends on whether or not its designers are familiar
with the principles of motion study in the field of IE.

In lune 1976, six months after we had started the Quality

Control Problem Study Group, we started the Motion and Time
Study Group, its tasks were to develop:
• Management by standard time
• Principles and techniques for operations improvements
• Skill analysis
Although the objects of study were different, the basic
principles for the study group were the same as those for the
Quality Control Problem Study Group.

II
112 MANAGERIAL ENGINEERING

Figure 5-1 An Example of Operation Analysis

 Reliable Method. IE Improvements 1 1

5.2 REDUCING SETUP TIME

At this time we were behind other firms in the reduction of

setup time. Therefore, as a baci<ground for our research, we vis-
ited several firms that are successful in this field. They very gra-
ciously spent time with us and discussed the problems they
had encountered and the results they had achieved. We often
heard: "We even worked on Saturdays and Sundays," or: "The
manager never went home before P.M. during that time."
1 1

Although each case seemed different from others, we

found a common factor in these successful cases. It was no-
thing but the analytical approach suggested by lE's
forefathers, about seventy years ago. We adopted this ap-
proach to their problem of setup time reduction with a single
modification that all concerned employees participate, rather
than an isolated handful of industrial engineers.
More approach suggests that operations
specifically, this
be divided into their basic elements, as is seen in Figure 5-1.
The importance of each of these elements is then weighed. The
basic steps for setup time reduction as developed by the Mo-
tion and Time Study Group are shown in Table 5-1. We com-
piled this approach from various successful cases in other
companies by analyzing them according to the perspective of
traditional IE.

5.3 THE "BUNT" STRATEGY

At the 'idea generation" step in Table 5- 1 , it is always possible
to find an alternative idea which achieves 70 to 80% of the ef-
fects atlO% of the original idea's cost. Engineering staffs tend
to conceive of ideas which may achieve impressive results, but
at an excessive cost. Making large investments also requires a
lot of time for decision-making. A plant that has only this kind
114 MANAGERIAL ENGINEERING

STEP DESCRIPTION
Choosing a setup Set the objective clearly
operation eg shorten the setup time by 80% for:
small lot production
reducing operators
expanding production capacity
Choose an operation which greatly needs
improvement
eg operations which;
take a lot of time
are very tiring
involve difficult adjustments
are very frequent
Choose a specific machine rather than a group of
machines
Observation and Measure average mean time in typical case (A
measurement small number of cases will suffice.)
eg cases of major products
cases of products with frequent production
Apply time study for short cycle operation
Apply operation analysis at one minute intervals
for long cycle operation, making description at
level of elemental operation.
Measure operations with higher frequencies at a
more detailed level.

Summing up Prepare a summary form and fill it out Post it where

it will attract suggestions from many people.

Analysis Examine the purpose of each elementary operation

and possible results if it is abolished Examine
from as many viewpoints as possible
Use cards —
they are effective in attracting and
presenting many peoples ideas without
requiring meetings.
Analysis by cards

Un.t

Ope
 Reliable Method: IE Improvements

STEP DESCRIPTION
While examining purposes, try to improve
operations that:
require a tiring posture
require physical strength
are unpleasant to perform
interrupt the rhythm of work
require attention

Idea generation
Hold meetings to gather ideas. Generate as
many ideas as possible which can be realized
quickly and cheaply even if their effects are
limited.
Utilize hints for reducing setup time:
( 1
) Draw a between internal setup
clear line
(done while a machine is stopped) and
external setup (done while a machine is
working).
(2) Change internal setup to external setup.
(3) Apply one-touch fixing and removing.
(4)Eliminate adjustments.
Develop alternative ideas with 80% effectiveness
and 10% cost.

Execution

Maintenance

Repetition
116 MANAGERIAL ENGINEERING

of improvement project characterized by long lead times will

generally be subject to low morale among those employees
who are most concerned by those improvements.
There is a surer way than this "home run" strategy: that is,
scoring by a series of hits and bunts. Starting in December
1977, six months after the beginning of the setup time reduc-
tion project, the Motion and Time Study Group began to ob-
tain the results shown in Table 5-2. The data were gathered
four times, each showing about 40% of the average rate of re-
duction. Figure 5-2 illustrates distributions of the reduction
rates, it should be noted that the proportion of projects with
rates of 70% or more remained constant throughout the three
investigations. The data show that by following the basic steps
30% of the projects will achieve reduction rates of 70% or more.
The study group achieved its initial objective, that is, the de-
velopment of a reliable method that can be used by anyone,
anywhere, to realize the expected results.
Table 5-3 gives an example of setup time reduction. The
change of dies that used to take sixty-one minutes and fifty-
five seconds now takes seven minutes and five seconds. The
number and bolts used was drastically reduced from
of nuts
and tools, from five to one.
fifty-eight to eleven,
Our approach to setup time reduction was a slight modifi-
 Reliable Method: IE Improvements 17

^ ^
118 MANAGERIAL ENGINEERING

Unit Operation
 Reliable Method. IE Improvements 1 19

cation of the basic IE approach. This modification included

microscopic examination of an operation and idea generation
by all the people who were concerned. It enabled us to:

• Easily eliminate purposeless motion

• Save improvement costs
• Obtain many new ideas
• Try these ideas quickly
Before the development of this approach, had once I

asked a plant engineer how he was proceeding (in terms of

Table 5-3). He answered: "Reducing the die setup time is our
main project. We have tried three ideas with considerable re-
sults in the past three years. Now we've got another plan, but it
will cost fifteen million yen, so we are trying to get budget ap-

proval. We hope to complete the plan in ten months."

The difference between the costs and speed of these im-
provement projects is staggering.
I remember a meeting where a three-year, long-range
plan was being discussed. Actually, there was little discussion.
But when the meeting was over and the topic of the layout of
the bicycle parking lots came up, people suddenly began to
talk with animation. Someone remarked that it was a pity that
the discussion of the long-range plan had not been so lively.

Rather than be sorry, it would make more sense to break

down the discussion of the long-range plan to the level of
"parking lots" and think about that problem in very basic
terms, if we only look at problems at the macro level, our situ-
ation appears unchanged over the past couple of years. The
same people discuss the same problems over and over, such
as weaknesses in quality control, subsidiary management, etc.
However, if we look at problems at a more detailed level,
we find many differences between today and three years ago.
We must be able to perceive such differences in our situational
and problem analysis. I will discuss the analysis of situations
in Chapter 7.
120 MANAGERIAL ENGINEERING

5.4 IMPROVING MAIN OPERATIONS

Returning to the setup time reduction projects, let me intro-

duce another important lesson we learned through our experi-
ences. Improving setup operations is nothing but improving
main operations; what we obtained in setup time reductions
can be applied to the improvement of main operations them-
selves.
Formerly, the most common goals of industrial en-
gineers, such as the minimization of idle time or the optimiza-
tion of layout, material handling, etc, did not always involve
the improvement of main operations. was not without cause
It

that workers often complained that their jobs became harder

after improvements.
Our approach to reducing setup time eliminates wasteful
and purposeless elements in main operations, and enables
the worker to do his job with much less time and effort. Thus it
comes as no surprise that our approach gained the full and en-
thusiastic support of workers.
In a previous chapter Ispoke of the evidence we possess
for actual work improvements. We obtained this skill through
experimentation in many projects and through observing hun-
dreds of cases. Anyone can apply this approach if he has
sufficient opportunity to experiment with it in real situations.
It does not require special talent. Consequently, we developed

a three day course for employees of Sumitomo Electric and

affiliated companies that would allow anyone to use this
method, (in addition to this IE course |Table 5-5|, we also de-
veloped courses in quality control and production control.)

5.5 ACHIEVING IMPROVEMENT THROUGH PRACTICE

The IE course was intended to teach the following lesson: In

spite of the numerous texts we have on IE, and the fact that
 Reliable Method. IE \mprovements 1

many people possess great knowledge of IE methods, there

still remains much room for improvement in blue collar as well

as white collar jobs, ifyou observe a job through analytical

eyes, it is not difficult to find waste in the way it is being done.
I suppose that many readers have had a similar experi-
ence, that is, applying the techniques of IE to analyze a prob-
lem, without a significant measure of success. When results do
not merit the time and effort we have invested, we often be-
come discouraged with them. In fact, it is not at all exceptional
to see two industrial engineers spend six months developing
an idea that will economize one worker and may never even be
put into practice. Since IE should be a tool for solving prob-
lems effectively and efficiently, it is no wonder that we become
quite discouraged with such experiences.
One of the important lessons that we learned is that the
cause of failure should not be attributed to the IE methods
themselves or to our knowledge and understanding of them,
but rather to insufficient practice. We must practice IE
methods enough so that we become capable of spotting op-
portunities for improvement at a glance.
It is often difficult, however, to have sufficient time to
and still perform one's own work. After
practice IE applications
having learned basic IE methods, we must make an effort to
practicethem within the limited time that is available to us.
The best way to do this is to observe daily operations around
us consciously and analytically.
For this observation,we must be sure to watch just which
parts of a worker's body move (Figure 5-3), and locate inef-
ficient motions by comparing them to the principles of motion
economy (Table 5-4). Such a conscious and analytical attitude
is called motion mind in the field of industrial engineering.

IE defines those who have motion mind in the following

three levels;
(I) Find differences among different motions (find and
detect problems).
122 MANAGERIAL ENGINEERING

Fingers 0.129 seconds

I Palms
I
- 0.258 seconds each
Wrists

Forearms 0.38 seconds

Upper arms 0.51 seconds

I

Shoulders
I

Waist - 0.645 seconds each

Walking

Figure 5-3 Body Motions and Average Time Required

(2) Decide which motion is preferable (explain the

reason).
(3)Develop a superiormotion (develop an innprovennent).
you practice observing motion (not only in the plant but
If

in the office as well) with motion mind, you will gradually be-
come able to detect inefficient motion quite easily. Such prac-
tice does not take much time. Once an inefficient motion is de-
tected, the desire will appear to eliminate it. This is the driving
force for improvement.
You should not tackle a large problem right after learning
a basic IE method. Rather, you should practice observing with
motion mind on a daily basis, finding and solving the prob-
lems you detect, however small they may be. Daily operations
provide a good opportunity for practicing IE methods this way.
Once you have developed the capability to achieve im-
provement through this practice, you can also help those
around you develop motion mind. Find inefficient elements in
their work, explain why these elements are inefficient, and
 Reliable Method: IE Improvements 1 23

show how to make the necessary improvements. It is impossi-

ble for foremen to spot every inefficient element in every
worker's motion even if they do possess excellent motion
mind. We cannot expect them alone to raise the overall pro-
ductivity of an entire plant.
on the other hand, all of the workers possess motion
if,

mind and try to improve their own work, improvements will be

much accelerated with remarkable results. No matter how
small each worker's individual role may be, a large population
with motion mind can achieve far larger effects on the entire
company than a limited number of distinguished engineers.
Figure 5-4 illustrates my idea about the difference.
it is not easy to increase labor productivity by 50%. I rec-
ommend to readers that they try to estimate how much invest-
ment is needed for 50% productivity growth. They might doubt
me if I say that we achieved it with little investment and with-
out forcing workers to work harder. The fact is that we did ob-
and motion mind. We
tain these impressive results through IE
must not overlook this approach and set off on a hasty path of
massive investments in robotization and automation. IE im-
provements must provide the basis for robots and automa-
tion.

5.6 POINTS ON THE OBSERVATION OF OPERATIONS

The following points on the observation of operations should

be noted:
All employees must understand the reasons for oper-
(
I
)

ations improvement. If one tries to measure and improve an

operation with a stopwatch or video tape recorder without giv-
ing an appropriate explanation to the people who are con-
cerned, he will obtain neither their participation nor accurate
data. It is indispensable to provide an adequate understand-
ing of the goals and purpose of operations improvement. The
124 MANAGERIAL ENGINEERING

improvement target should be determined through total par-

ticipation.
The observed worker must understand the purpose of
(2)
observation. Most workers are unable to relax while their oper-
ations are being measured. They may also misunderstand the
reason they are being observed. once witnessed a case where
1

a worker was too anxious to continue his soldering work be-

cause he had not been given any explanation beforehand. felt I

sorry for him.

It is essential to provide a satisfactory explanation for ob-
servation and measurement in order to have the workers con-
tinue to work as usual. Operations observation and time mea-
surement are conducted to find time differences between dif-
ferent methods, not to compare operation speeds between
workers.
However hard a worker works, he will be unable to raise
productivity significantly if his work method involves many in-

efficient elements. On the contrary, high productivity is possi-

ble ifwork method is itself efficient. Although
a to it is difficult

measure accurately the ease and efficiency of an operation,

the time needed to perform it can be one reliable measure.
(3) An actual operation should be observed in detail. We
do not necessarily know every even in our
detail of operations,
own plant. This is because we do not have actual work experi-
ence, or different workers perform the same operation differ-
ently. In the latter case, we must determine the reason for
these differences. Workers know about their methods for per-
forming different operations better than foremen in most
cases.
As seen in the cases of CEDAC applications, we often find
that workers' methods often deviate widely from standard pro-
cedures which maybe impossible to follow in real work condi-
tions Or one may observe that workers take more time than
had been expected to adjust old, worn-out equipment. In such
 Reliable Method. IE \mprovemenls 125

Improvement
level

Figure 5-4 Accumulation of Small Improvements by Large Workforce

cases we must solve problems that obstruct efficient opera-

tions. For this purpose too, we must observe every detail of
current operations first.

Detailed observation is easy, but seldom practiced.

Analyzing elemental operations is good practice in observa-
tion. Even a little practice is enough for mastering basic oper-
ations analysis.
These detailed observations must be made for at least
thirty minutes, in this process, many items requiring improve-
ment are detected. Although these may turn out to be minor
improvements, their aggregate effects are considerable. Ele-
mental operations analysis is discussed in most textbooks on
IE. Those readers interested in pursuing this subject can con-

sult any one of them.

5.7 PRINCIPLES FOR THE IMPROVEMENT OF

OPERATIONS

One often hears such statements as: "Recent order sizes are
decreasing, and automation is accordingly becoming more
126 MANAGERIAL ENGINEERING

and more difficult. Furthermore, since each product is de-

signed and manufactured differently, it's difficult to improve
overall productivity." However, if one observes operations in
these plants carefully, he some common elemental
will find

operations leaving considerable room for improvement. Prin-

ciples for discovering such elements are introduced below.
Principle 1 . Use the principles of motion economy to detect

wasteful motions.
The larger a part of the human body, the longer it takes to
move. Therefore, the most efficient operations involve move-
ments of fingers only (Figure 5-3). Such ideas are systematized
in the principles of motion economy shown in Table 5-4. These

principles are useful in finding wasteful elements in human

motions.
One may feel, when observing a production job, that "it's

tiring," "it's slow," or "it doesn't have any rhythm." In such

cases, it is possible to determine what elements are wasteful
by consulting these principles. This simple method will help
eliminate wasteful elements in many cases.
The principles of motion economy also enable us to ob-
serve operations in a new way, that is, to compare them with
ideal conditions. It is not so difficult to imagine ideal opera-
tions. Consider an assembly operation, for instance. One
should be able to pick up the necessary parts quickly and with
a minimum of movement. The assembly job should not in-
volve checking part types, assembly positions, adjustments or
repairs. The worker must be in a natural posture, moving his
arms and legs with good balance. When the assembly has
been completed, the goods should be transferred to the next
process within a minimum distance. In other words, the ideal
operation involves only those elements that add value to the
product. All other elements should be eliminated. Comparing
the ideal operation with the present method, one may easily
detect which aspects of an operation should be improved.
 Reliable Method. IE Improvements 127

V)
JJ
flj

CQ
28 MANAGERIAL ENGINEERING

Case in the appendix shows how an operation was improved

1

by applying the principles of motion economy.

Principle 2. Analyze the necessity of the operation.

Many operations considered necessary can be in fact

simplified or even eliminated if they are approached from a
different viewpoint. Detailed and analytical observation is es-
sential for this purpose. Areas for improvement are detected
through the process of analyzing each elemental operation's
necessity and objective. Ideas obtained in this way often elimi-
nate an entire operation.
Analyzing necessity is straightforward enough. One has
only to ask a series of questions such as: "Why do I perform
this operation?" "Why do do it in this order?" "Why do have
I I

to check it?" The answers to these questions determine the

necessity of a given operation.
Such questioning is often proposed in plants, but most
people stop at this stage and obtain only limited results. The
next step is more important: to develop many alternative ways
of reaching the same objective. For example, "Can include it 1

in the previous process?" "Is there another way can do it with- I

out adjustment?" "Can eliminate it completely?" and so on.

I

If one attacks a large operation as a whole, he will find

that it includes various necessary elements and is very hard to

improve. Dividing it down to the level of elemental operations

or even unit movements isthekeytogeneratingan abundance

of new improvement Such a process can produce as-
ideas.
tonishing ideas for improvement (see Case 2 in the Appendix
to this chapter).
Principle 3. Observe the previous and subsequent processes

together.

Principles I and 2 aim at improvement in a single process.

However, many improvements may be secured by observing
subsequent processes at the same time. This type of improve-
ment has a greater effect in most cases. Clues for this type of
 Reliable Method: IE Improvements 129

improvement include:
• The placing and picking up of products are done
separately between the two processes
The two processes can be integrated
The two processes can be divided differently
One worker can be assigned to the two processes
An operation can be transferred to a different process
The work can be divided differently between workers so
that all the workers have the same cycle time. This often
requires changing equipment layouts, as seen in Case 3.

Improvements are often made by means of these con-

cepts. For example, a delivered product is dismantled for the
inspection of inner parts, reassembled and then taken apart
for other checks, etc.Extreme division of labor often causes
such wasteful operations.
It should be noted that most advanced machine tools can

be used to perform a variety of operations once they hold a

product. Picking up and placing operations must be kept to a
minimum. They are wasteful operations and tend to exacer-
bate unnecessary costs.

5.8 AN INVITATION TO IE IMPROVEMENT

The reader may suspect that the method introduced above is
too old-fashioned for today's rapidly changing world. He
might prefer more drastic measures suitable to this age of au-
tomation. A powerful medicine may be effective in the short
run, but it cannot create lasting physical strength. do not I

deny the necessity of such medicine in certain cases, but am I

afraid that people, longing for such medicine, overlook many

improvement opportunities under their noses.
right
briefly introduced a method which can give power to
I

people for solving problems on the job. If a company obtains

130 MANAGERIAL ENGINEERING

such improvement power, it can grow anywhere, no matter

what problems it must surmount. If its competitive advantage
lies only in equipment, its competitors can easily catch up

with by making the same investment. No company can rely

it

solely on
capital investment to improve productivity, if it pos-
sesses some strength which takes a long time to develop, its ri-
vals will need an equally long period. Such strength can pro-
vide the basis of a long-lasting competitive advantage.
There another important point, A new technique, no
is

matjter how old-fashioned it may appear, is nothing less than

an engineering improvement if it allows us to increase produc-
tivity by 50% without making large capital investments or in-

creasing work intensity. If you obtain improvement power

through the method described above, you will find similar
seeds for further engineering development in your own work.

5.9 THE CONTRIBUTION OF WORKER GROUP

ACTIVITIES

Once there was a time when terms such as "modern IE" or "ad-
vanced IE" were very popular. To differentiate between them,
people coined a funny term, "traditional IE." Since science and
engineering keep advancing, the adjectives "modern" and "ad-
vanced" really do not make much sense. Whatever the value of
this term, the concept will introduce below belongs to the
I

realm of traditional IE.

It also concerns the QC circle activities of our Japanese

workers which are now world-famous. wonder if the reader
I

knows that two-thirds of the improvement proposals made by

worker groups are developed by traditional IE methods. This
chapter introduces an engineering approach to the activities
of worker groups.
The remarkable results observed in so many Japanese
firms are attributed to those basic traditional IE methods
 3

Reliable Method: IE Improvements 1

rather than to group activities. The latter constitute a medium

for the application of IE methods. This idea is illustrated in
Table Although production engineers may not be as in-
5-5.

terested in traditional IE methods as in robots and micro-

computers, in fact they are the actual driving force behind our
productivity growth.
On would be a mistake to overlook the
the other hand, it

impact of worker group activities on human relations. Com-

munication and cooperation between workers and managers
are improved, and opportunities for self-improvement are pro-
moted.
Returning to productivity, we wanted to determine quan-
titatively how much of a contribution worker groups made re-
lative to the contributions of our engineering staff. We used a
statisticalmethod called multiple regression analysis to ex-
plain man-hour efficiency growth by worker group activities
(represented by the number of improvement suggestions per
man-year). The Production Engineering Department com-
pared this to efficiency growth, which was measured by capital
investment in equipment. Data were collected from nine
plants over a period of three years. Figure 5-5 shows this
model.
The number of proposalsannotated with "as a repre-
is

sentative of the aggregate measure," because it represents an

aggregate measure which was developed out of twenty-seven

variables concerning the plant's management and improve-
ment power. These variables include quantitative factors such
as the number of proposals, hours of group meetings, hours
spent on worker education and training. They also included
qualitative factors such as how group activities were activated,
how we followed up on group projects and how workers were
motivated. The aggregate measure developed from these vari-
ables showed the highest correlation with the number of prop-
osals, so the latter was chosen to represent the former as an
easily understandable measure.
32 MANAGERIAL ENGINEERING

><

QQ
o
 Reliable Method. IE \mprovements 1 33

Y = 8.46 \Ix7+ 14.66 vlx^- 2.78

Y = MH efficiency growth (%/year) (or, annual average MH

efficiency of the year less that of the previous year)

Xi = Number of improvement proposals per man-year as a

representative of the aggregate measure

X2 = Capital investment on equipment ($1000/man-year)

Figure 5-5 A Regression Model for Efficiency Growth

The graphed in Figure 5-6. in the

results of Figure 5-5 are
figure, arrow A shows the additional anaount of capital invest-
ment required to raise annual productivity growth from 20 to
25%. Arrow B shows the number of proposals needed to secure
the same effect. Rather than depend only on one or the other,
a combination of both, shown as arrow C, proves to be the
most effective means of boosting productivity.
Another point this figure shows is that one proposal has
the equivalent effect on productivity growth as a $3,000 capital
investment. However, we found that most improvement sug-
gestions cannot produce such a significant improvement ef-

fect. We analyzed the question as follows.

As Figure 5-7 shows, the true contribution of worker
groups to efficiency comes from the combination of their im-
provements and day-to-day management.
Day-to-day management is a system that supports worker
group activities. As seen in Figure 5-7, standard times are set
for current work methods, and man-hour efficiency is mea-
34 MANAGERIAL ENGINEERING

c
 Reliable Method. IE Improvements 135

r
136 MANAGERIAL ENGINEERING

In day-to-day management, efficiency

is measured on a

Problems which appeared yesterday are detected

daily basis.
today, and appropriate countermeasures are taken without
delay. Also, production status compared with the plan,
is

which may call for corrective measures as well. This day-to-day

management system is making a significant contribution to
productivity improvement in many plants.
The regression model in Figure 5-5 indicates quantita-
tively the contributions of worker group activities as well as
production engineering improvements. We
should note the
importance of the first term — improvement
proposals —
which represents improvement activities by worker groups.
The model predicts that these groups can significantly con-
tribute to increased productivity in any kind of plant, domestic
or foreign. The coefficient that stands for the degree of con-
tribution will of course differ from plant to plant.
The number of proposals is a representative measure of
the plant's overall level of management and improvement
power. If a plant tries to increase only the number of propos-
als, neglecting other things, cannot expect efficiency growth
it

as predicted by the model. The number of proposals must be

increased while raising the level of management and improve-
ment power The fact that there is a high correlation between
the number of proposals and factors of management and im-
provement power implies that the latter stimulates the former.
In other words, plants with a high level of management and
improvement power can expect more improvement proposals
from worker group activities.
Concerning the relation that a higher level of manage-
ment and improvement power leads to a larger number of sug-
gestions, there is an important point that was mentioned in
the previous sections: If one has a reliable method for detect-

ing every improvement opportunity, and practices that

method sufficiently, then he can develop far more improve-
ment proposals than other people. Therefore it is no wonder
 Reliable Method. IE Improvements 137

that higher improvement power leads to more improvement

suggestions. And, as have said in previous sections, we do
I

possess a reliable method, IE, which gives improvement

power to worker groups.
if all the workers in a plant attain this improvement
power, it will become a truly amazing organization, capable of
achieving tremendous victories in the quest for productivity
improvement.

APPENDIX TO CHAPTER 5
THREE CASE STUDIES

Case I

A stamping operation in which both sides of a part are

stamped.
Operation prior to improvement:
(1) Place a pallet at point A.
(2) Get parts from the pallet and place them at B.

(3) Get a part at B and stamp on its upper face.

(4) Put the part at C.
(5) Turn the part and transfer it to B.
(6) Get a part from B and stamp its lower face.

(7) Put the part at C.

(8) Transfer parts from C to the pallet.
This operation involves unnecessary steps, namely, (2),

(5) and (8).

Press
138 MANAGERIAL ENGINEERING

Operation after improvement:

{ 1) Place a pallet at A. (Use rimless pallets to facilitate
picking up).
(2) Get a part from the pallet and stamp upper face.
(3) Put the stamped part at B (Finish all the parts.)
(4) Get a part from B: turn and stamp it.
(5) Put the part on the pallet at A.
Having eliminated unnecessary placing and picking mo-
tions, efficiency was improved by 50%.

Press

Case 2
A material cutting operation which material for silver con-
in

tacts is cut with a power cutter. The worker sandpapers an edge

of material after each cutting. He was asked the following
questions:
Q: Why do you sandpaper the material?
A: To remove scrap.
Q: Why do you remove scrap?
A: If there is scrap at the edge. I cannot cut the material
vertically.
Material is cut with a cutter and guideboard as shown on
the left panel. If cutting scrap is not removed, the material can-
not be positioned horizontally and cut at the correct angle. An
improvement was obtained by making a small gutter on the
guide board so as to allow correct positioning of material in
spite of the rough edge. By eliminating sandpapering, the av-
 Reliable Method: IE \mprovements 39

erage cycle time of .34 minutes was reduced by 50% to .17

minutes.

A ^ Cutter

Cutting scrap

ir y ^
Material
^

—
A guide
^
board
mr
A gutter

Case 3
A drum-rim repair operation. After electric cables are installed,
their drums and returned to the manufacturer.
are dismantled
Before re-use, these drums are repaired if any flaw or distor-
tion is detected on inspection.

Operation prior to improvement:

( 1) Assemble a drum.
(2) The assembly man inspects the drum by turning it
and decides if it needs repair.
(3) Non-defective drums are sent to the paint shop.
(4) Defective drums are transferred to the distortion
repair machine.
The repairman turns the defective drum, detects
distortion and repairs it.
(6) drums are sent to the paint shop.
After repair,
in this operation, the transfer and inspection at (4) and (5)

are pointless. The same inspection is done twice by two

people. After improvement, the repair machine was moved to
the assembly worker so that he could repair defective drums
right after inspection.
chapter 6

TWO APPROACHES TO
"PRACTICE BY ALL"

6.1 OET: UTILIZING ERRORS FOR PRACTICE

Only when everyone has mastered a reliable method through

constant practice can permanent positive results be attained.
In the case introduced below, a system called OET (On-Error
Training) was developed as a way of training workers to learn
from their errors. The underlying principle of OET is that work-
ers can best learn to improve quality and productivity by
studying their mistakes and systematically practicing a
method for correcting them.
The forming section of the plant that developed OET pro-
duces 1,300 types (180,000 units) of sintered alloy parts each
month. Setup operations in this section occur as many as
1 3,000 times a month. Semi-finished products before sintering
are as brittle as chalk, so many chips and cracks appear in
them. Also, since products shrink in the sintering process, size
abnormalities are often observed. It is necessary for the work-
ers in this plant to have the skill to decide appropriate
methods and orders of processing as well as the skill to handle
products very carefully.
Because of these difficulties, many defects due to care-
less work were produced —
even by veteran workers. The de-
fects included incorrect sizes, angles and shapes, as well as

141
142 MANAGERIAL ENGINEERING

100
Application of CEDAC

80

^. 60

CO
(Education in the group

o .Exhibition of defective samples

o
<x>
.Introduction of OET
40
Q
>s
Q)
Group discussions on how to
C ^minimize human error
o
|OET at the place defects appear
c
z> 20 Re-education
o I

E
< Re-education
10

1976 1977 1978 1979 1980

Figure 6-1 Trend of Work-In-Process Quality Defects

 Practice By Air 143

chips and cracks, caused by a lack of basic knowledge such as

the ability to read drawings or use simple measurement tools.
This resulted from inadequate on-the-job training and weak-
nesses in operations standards.
mid- 1976, CEDAC was introduced to this plant along
in

with programs designed to provide basic knowledge and on-

the-job training. As a result, losses due to defects were cut by
two-thirds (Figure 6- This level continued through the latter
1
).

halfofl977, and was considered the minimum level attainable.

Nonetheless, the repeated appearance of defects indi-
cated that there was still room for improvement in communi-
cations, training, standards and morale. To analyze the true
causes of these defects, the New loharry's Window described
in Chapter 3 was used. Figure 6-2 was developed by four group

leaders and several veteran workers. They analyzed past defect

records issued during January and February of 1978, and class-
ifiedeach defect into one of four categories given in the
window.
It turned out that there were more defects in categories 11
and III than in IV, which indicated that the causes of most de-
fects were known to some employees. Some workers knew how
to do their jobs correctly, but this knowledge was not
adequately communicated to the others who thus continued
to make mistakes.
There are several reasons why a worker may not correctly
perform an operation that he knows well. He may make care-
less or human errors. He may not have sufficient skill to follow
the correct operations. He may even consciously deviate from
them.
In addition to analyzing these problems with the New
loharry's Window, workers were trained in the basic knowledge
they needed to perform their jobs correctly. Starting inMay
1978, defects which cropped up were exhibited for a month to
increase the workers' awareness of them. This helped reduce
144 MANAGERIAL ENGINEERING

\Member
 "Practice Bi^ AH" 145

defects considerably, but they were still not completely elimi-

nated.
At this time the workers were divided into two groups and
had many heated discussions on whether carelessness, which
led to errors in reading drawings and choosing correct tools,
could be completely eliminated. One group insisted that it
was impossible. The other believed that careless errors could
be removed if their causes were eliminated. The OET activities
were instituted to focus this heightened awareness on defect
reduction. Our goal was to have all the workers think and teach
each other as soon as any defect was detected. We intended to
utilize defects as good opportunities to teach and practice cor-
rect operations.
Five rules were established for on-error training:
The quickly rule
(1)

(2) The actually rule

(3) The himself rule

(4) The don' speak rule

t

(5) The support rule

The quickly rule requires that the person who caused the
defect call all the group members within thirty minutes to dis-
cuss countermeasures. This meeting would not last for more
than thirty minutes.
The actually person who caused the de-
rule requires the
back the operation, exactly as he had done it, to the
fect to play
group. Causes and countermeasures are also discussed in the
same place. This makes it easier for the others to understand
how and why the defect occurred.
The himself rule requires the person who caused the
defect to explain it to the others himself.
The don't speak rule prohibits foremen and group leaders
from speaking first. It aims at giving other people a chance to
speak, because if the foremen or leaders present their ideas
first, other members may not think by themselves and may feel
inhibited.
 .

146 MANAGERIAL ENGINEERING

Foremen and leaders are only allowed to speak at the

final stage of the nneeting. They are expected to support group
activities behind the scene. This is called the support rule.
As a and improvennent
result of these rules, discussions
by the worker groups came about more frequently than we had
originally expected. Figure 6-3 shows the result of OET after it
had been running for four months. Compared with the data of
Figure 6-2, it is easily seen that defects had been drastically re-
duced.
In ordinary worker training programs, workers tend to be
passive, in OET, on the other hand, they are expected to ac-
tively teach each other. Thisis the reason we use the word

"practice," which has an active connotation, rather than "train-

ing," which suggests passivity.
OET is a method for building quality control activities out
ofdaily work, and it suggests a new direction for QC. its four es-
sentialelements are shown in Figure 6-4, and it is compared to
ordinary group activities in Table 6- 1

(2) CONCERN
(1) BASIC KNOWLEDGE Rotate jobs so that people can
Understand defects and learn show concern about other
about countermeasures. shops' defects.

(3) COMMUNICATION OF
^5^ (4) ADHERENCE TO
INFORMATION ABOUT DEFECTS ESTABLISHED
AND DEVELOPMENT OF MEASURES (CEDAC
EFFECTIVE COUNTERMEASURES and OET)
(CEDAC and OET) Use measures that can
Transmit information on defects and actually be put into
countermeasures promptly and practice.
accurately. Provide environment
Give all people concerned complete that facilitates
information. adherence to
Develop workable measures from all measures.
the people's ideas and experiences.

Figure 6-4 Four Essential Factors for OET

"Practice By All" 147
148 MANAGERIAL ENGINEERING

A worker once said to me: "We were afraid we were spend-

ing too much time on OET. At the outset, we spent 190 man-
hours per month, which equalled 3% of the section's total
man-hours. But as we took effective measures one after the
other, man-hours for OET decreased; the present level is 40.5
man-hours, or .6% of the total."
Thus we learned the importance of investing in education
and practice from the OET experience.

6.2 PRACTICING QC BY THE CASE METHOD

Many, many OC methods have been developed, and they are
taught in many seminars and textbooks in Japan. Unfortu-
nately, these methods are not often used effectively in real
manufacturing situations. The importance of seminars and
textbooks for teaching QC cannot be over-emphasized, but we
also need efficient methods for practicing it. The best way to
practice OC methods is to apply them to actual problems.
However, most people, without good coaching, do not have
the wherewithal to go through this long trial-and-error pro-
cess.
To achieve a breakthrough, we developed a case method
for OC. Participants in this case method are expected to apply
their OC knowledge and methods to a fictional problem, and
to learn how to correct their weaknesses by studying their fail-
ures and listening to the instructor's advice.
This case methodadopted in a three-day OC program
is

for staff and managers within the company. About 300 people
participated in this program. They were divided into sixteen
classes, each with fewer than twenty members. The program
questionnaire we developed for participants showed that the
case method was ranked among the most popular courses. We
often heard comments such as: "It was very helpful. 'My con- "

fidence has been restored." "Now understand why our former

I
 "Practice By All" 149

QC activitieswere not fruitful."

Encouraged by this result, the third session of the mana-
gers' QC Convention of the Japanese Union of Scientists and
Engineers (where 1 serve as chairnnan) is now studying
techniques developing good practice cases.
for
would like to comment here on the case method. As is
I

widely known, the case method originated at Harvard Busi-

ness School as a unique approach to business education.
Wallace B. Donham, who was dean of the business school for
twenty-four years, laid the foundations of the case method. All
the courses at the school are taught according to this method,
which has earned renown as a landmark innovation in educa-
tion. Sincehave never witnessed the Harvard method in ac-
1

tion, lam not certain whether our own method differs from it.
it has often been noted that learning only from lectures

does little to prepare us for the practical application of that

knowledge in real situations. Once a new graduate joined a
company and was taken aback when his boss said to him:
"What can you do for us?" Since we spend a lot of time obtain-
ing knowledge at school as well as in the company, we expect
good results that justify our efforts.
One might be tempted good textbooks con-
to reply that
tain questions and exercises for learning QC. do not deny the I

necessity of exercises, but they differ from the case method on

the follov/ing points: The exercise question provides all the in-
formation necessary to solve it, and it has only one right solu-
tion. What is more, most of the examinations given in schools
have only one correct solution (at least from the teacher's
point of view).
On method does not give all the
the other hand, our case
information, and there may be more than one solution. In fact,
the more knowledge and information one has, the more solu-
tions he may discover.
In addition, solutions to case method problems can be
assigned priorities according to their effectiveness. The
 "

150 MANAGERIAL ENGINEERING

reasons behind these priorities must be very clear if they are to

persuade case participants. Finding as many solutions as pos-
sible and assigning them reasonable priorities are necessary
conditions for securing good results, not only in QC but in all

other fields.
Few problems in business provide all the necessary
real
information. Thus we are always constrained to making deci-
sions in a context of uncertainty. We have to estimate un-
known factors to apply basic decision-making models.
In the three-day program mentioned above, teams of four
to five people are requested to write out their judgments and
countermeasures for a case during the morning of the first day.
From then on to the afternoon of the last day. the group listens
to lectures. Then the same case is tried again by the same
people. They are expected to see how different their judg-
ments and proposals are following the lectures. Since the par-
ticipants are all able managers and staff, their proposals on
the first day are not necessarily poor, but if they are able to
note a difference between their first and second tries, however
slight, they will recognize the effectiveness of the education.
One particular suggestion from the instructor was well re-

ceived by all the participants: "In the past you have applied QC
methods at the level of the first trial. If you observed an im-
provement after the lectures, then you had better practice QC
from now on. This difference can be capitalized
at that level
upon as your competitive advantage over rival companies.
The difference between the first and second trial is due to:
• Lack of basic knowledge, or insufficient practice for apply-
ing the knowledge that a participant did possess
• Jumping to vague conclusions or proposals without
rigorously analyzing the causes
• Over-generalization
• Neglect of an action that should and can be taken
immediately, possibly because of over-generalization
 '

"Practice By All 151

• Lack of understanding of the importance of mathematical

techniques in QC
• Blind beHef in amateurish approaches to QC
Since there were abundant opportunities to learn QC in

lapan, mere enough to attract the interest of

lectures were not
the participants. They already knew most of the methods and
theories. Therefore, the case method was employed to help
them realize the critical importance of knowing reliable
methods and mastering them through practice.
The more competent one is, the fewer opportunities he
has to have his weaknesses in QC pointed out to him in his
daily work. Highly able staff and managers can increase their
QC skill by using the case method.
levels best
Iobserved an interesting fact through the case method.
That is, competent people have a very clear perception of their
weaknesses. Qn the other hand, those with insufficient knowl-
edge and ability do not readily learn from their failures. There
is an obvious causal relationship between one's ability and his

sensitivity to feedback. Those with an effective feedback "cir-

cuit" can learn from failures, improve their weaknesses, and
correct their mistakes, thus becoming increasingly able. ex- I

pect some participants learned something from reflecting on

their past failures and weaknesses in QC. Such a profound les-
son can be learned only from within.
1also employ a case method for the final examination of
the reliability engineering course teach at the Department of
1

Systems Engineering at Kobe University. have been lecturing I

at this school with its cheerful, liberal character and beautiful

mountainside campus since 1975. And it has been a great
pleasure for me to work with such dedicated students.
The final examination begins like this:
You are a reliability engineer hired by F.R. Bisman. an au-
tomobile engine manufacturer based in W^st Germany. You
belong to the Production Engineering Department of the third
152 MANAGERIAL ENGINEERING

plant. The foreman of the inspection process and the chief of the

final engine assembly shop reported a problem as described in

the attached sheet. They asked the plant manager what meas-
ures should be taken. The manager wants advice from you as a
reliability engineer. Describe your proposals briefly. This is

your first job at this company. I hope you do a good job.

The fact that a German company is used is not important. I

was just afraid that the students would be influenced by their

biases if it were a Japanese firm.
As a result of this preparation, have full confidence that
1

my students will be able to do a good job when their future

bosses ask them: "You took a course in reliability engineering;
what can you do to improve this product's reliability?"
chapter 7

SITUATIONAL ANALYSIS:
FIND YOUR PROBLEMS
YOURSELF

In Chapter described a case in which a situation favorable to

1 1

innprovements was developed by dennonstrating the effective-

ness of IE methods. The existence of reliable naethods as well
as sufficient practice of thenn is innportant in the developnnent
of such a favorable situation.
To create a favorable environment, correct situational
analysis is indispensable.

7. 1 CARTOONS FOR EOUIPMENT MAINTENANCE

Production equipment cannot be used to its full effectiveness
without the close cooperation of machine operators and
maintenance workers. In heavy industries, this equipment ac-
tually produces the products, so equipment maintenance is
the most important factor in production engineering.
When was manager of the equipment development and
I

maintenance section of an alloy products plant, tried to

I

develop a Total Preventive Maintenance (TPM) system in

which machine operators and maintenance workers work to-

153
154 MANAGERIAL ENGINEERING

gether. I first interviewed a maintenance worker. He said: "We

are machine repairmen, but are the operators machine de-
stroyers?" Human between operators and mainte-
relations
nance worlcers were not very friendly. In this situation, it was
not a good strategy to strengthen only one party. wanted to I

improve both groups' skills and knowledge at the same time.

The first thing did was ask the maintenance workers to
1

draw cartoons to express what they wanted the operators to

do. Fortunately, there was a very good artist among them who
drew twenty poster-size cartoons. still have several of them
I

now, nine years later. For example, one cartoon presents the
machine as a child, with the operator as its mother. The
mother brought the child to the doctor (the maintenance
worker), complaining that the child had a headache. But the
child's sickness was actually in its abdomen, and the doctor,
having been misled by the mother, had to go to great lengths
to find the true sickness. Other cartoons illustrated machines
crying because of lack of oiling, and a drive shaft worn out by
dust.
The cartoons were exhibited in the plant. After that, 1 sum-
marized everyone's ideas in the form shown in Figure 7-1. (Al-

though this form resembles Professor Kawakita's K| method, it

did not follow the same formal steps.)

7.2 REPLACING FORMAL MEETINGS

Figure 7- was drawn as follows: After exhibiting the cartoons,

1

I asked the operators and maintenance workers (who at that

time belonged to a different section), to write down their own
ideas for improved equipment maintenance. They were ex-
pected to write down one suggestion per card, and the number
of cards was unlimited.
A few days later, many cards were collected. I read them
 Situational Analysis 155

over again and again. By rereading, 1 was able to grasp the situ-
ation which was explained in diverse ways on the cards.
1 sorted the cards nto several groups and gave each group
i

a short title. The cards were posted on a large chart where

everyone could read thenn. I then asked the workers to let me
know had misunderstood what they meant on the cards.
if 1 I

proceeded to revise the titles until they were accepted by

everyone.
Some examples of the cards are given in Figure 7-1 but , all

the cards were posted on the original diagram.

7.3 CLARIFYING THE PROBLEM SITUATION

There were three basic groups of cards. The upper left group in-

cluded many cards indicating that what was needed for better
equipment maintenance ^Nas not skill but rather knowledge and in-
formation. The maintenance workers wanted better informa-
tion on machine breakdowns, while the machine operators de-
manded better explanations of the repairs. The upper right
group (on the chart) concerned equipment design-, since some of
the machines were poorly designed, they often broke down
and needed frequent repairs. The groups at the bottom in-
volved items on equipment improvements. As a machine wears
out, it needs not only repairs to restore it to its original condi-
tion, but also improvements so that it will be subject to fewer
repairs in the future.
After having thus analyzed the situation, I called all the
concerned and line workers together for a meeting to dis-
staff
cuss concrete measures for better equipment maintenance
with the full cooperation of the machine operators and
maintenance workers. realized at this time that we had usu-
1

ally held similar meetings without the appropriate situational

analysis.
56 MANAGERIAL ENGINEERING

EQUIPMENT DEMANDS KNOWLEDGE RATHER THAN SKILL (TEACH

MAINTENANCE IN MORE INTERESTING
WAYS)
— Give detailed explanation on breakdowns to
maintenance workers
Operators should have more knowledge of and
interest in equipment

EXPLAIN THE CONTENT OF BREAKDOWNS WITH

u CORRECT INFORMATION
Don cover up mishandling
t

of operations

ELIMINATING DUST IS THE

FIRST STEP
DEVELOP PREVENTIVE MEASURES ONE AT A TIME OFPM
-Pipes too narrow for high pressure Clean equip-
water cause breakdowns ment every
day
'Excessive fluctuation of rotary
pumps
'Heat molding machine needs
spare coils
SAFETY DEVICES
Check safety
I I [ I

devices before
each shift
I I I I

'

CONTROL OF LUBRI-
1 GATING OILS
I 1 I

^ Indicate the

I I I [
name of the
correct oil

where oiled

CLEAN WATER AND VACUUM PUMP

CLEAN INEFFECTIVE
I—T-lII
I
I
1

EQUIPMENT
1
I
——1

'
I

'
1

* -Vacuum pump
PARTS INVENTORY
I
'
I
'
I

'
I
'
Cooling water
too dirty
c5 I 1 filthy with oil
r 1 I
1
CONTROL
rSufficieni in-

ventory of
parts needed
for specific
breakdowns

Figure 7-1 Author's Sketch of Some Ideas on TPM Developed by

 1

Situational Analysis 57

NUMBER OF BREAKDOWNS SET EQUIPMENT STANDARDS

IS MEASURE ''Of^ALL
^
1
I
11 1

I 1 OFPM I

t— I 1 Gather data on
|- Inspect equip-
j
'
' ment often
(operators &
maintenance
workers)

QC AUDIT LEADING TO MAINTENANCE AND DESIGN

,
,
,
^EQUIPMENT 1
r'^o'' machines
I

[audit I l frequently
I 1

1
Equipment de-
Clarify true problems
I
—»— I I r^
|

breaking down.
I

I 1 signers must I I I I preventive

check their maintenance by
equipment design mainte-
in operation nance sections
58 MANAGERIAL ENGINEERING

Former meetings had always been opened with: "I would

like you to tell me your ideas on how to improve our total pre-
ventive maintenance system." Since the problem situation
was not clearly defined in the minds of the participants, the
meetings tended to zig-zag: Mr. A would talk about the work-
ers' lack of knowledge of the equipment while Mr. B expressed

the same idea using different words. Mr. C would then inter-
rupt to talk about the breakdown of a vacuum pump, and so
on. Since no one grasped the total situation, the discussion
could not be focused. The presentation of ideas was not well
prepared, so one was constantly subjected to useless intro-
ductions, repetition and changing topics. One person might
respond before understanding exactly what the other was talk-
ing about. The fruitless discussion would continue inde-
finitely...

doubt that such meetings truly deserve the name of

1

"conference." To adequately grasp the problem situation, a

simple method as in Figure 7-1 is sufficient. If a meeting is
called expressly to collect this amount of information, it will

take great time and effort.

Since that time,have been using this simple method for

1

situational analysis instead of conferences. learned in an ex-

I

perience which will describe below that this method would

I

correctly resolve our problems and that it is a powerful tool for

situational analysis.

7.4 DETECT YOUR OWN PROBLEMS

At this time, SumitomoElectric had a technical audit system,

and our equipment maintenance was inspected by five
plant's
veteran auditors a year and a half before was assigned to the
I

section. After the situational analysis had been conducted, 1

came across their report and was surprised to read that those
auditors, using their very powerful techniques, had reached
 .

Situational Analysis 1 59

exactly the same conclusion 1 had by my simple method.

1 learned from this experience that the method sufficed to
detect problems and that we therefore needed no external au-
ditors. When 1 was the manager of an IE department that in-

cluded technical audits among its many tasks, several direc-

tors asked us to inspect them, instead of accepting their invita-
tion, always encouraged them to conduct a situational
1

analysis by themselves, it is almost impossible for external au-

ditors to analyze a situation better than those who actually
face it. based on the assumption
This is that internal people
do understand their own situation.
their best to
believe that situational analysis by people who are actu-
1

ally concerned is the best way to gather and integrate informa-

tion correctly. And it is a powerful advantage to have a reliable
method for conducting such situational analysis swiftly and
correctly.
in the case of the alloy products plant, a meeting was
called only when the situational analysis had been completed.
The machine operators and maintenance workers were asked
to develop improvement ideas based on the results of the
analysis. From this analysis, it had become clear that the
operators required more knowledge of the equipment. Some
kind of educational program had to be developed.

7.5 CORRECT SITUATIONAL ANALYSIS AND

CORRECT ACTIONS
We decided to take the following measures;
The foremen for the two maintenance groups, one for
machines, investigated break-
electric facilities, the other for
downs during the previous year that had been caused by in-
sufficient worker knowledge. These covered 23% of the total
machine breakdowns. They analyzed these breakdowns and
160 MANAGERIAL ENGINEERING

wrote a short textbook for operators with basic but indispens-

able knowledge on equipment.
They told the operators:
"Twenty-three percent of last year's breakdowns were
caused because you don't know your machines very well. We
have collected very important knowledge in this notebook that
we want you all to know. If you master this textbook, however
hard it may be, you will be able to reduce breakdowns by 23%.
Therefore, please do your best to read it."
The era of skilled veterans is over, and operators are now
required to have appropriate knowledge to use sophisticated
equipment. The foremen offered no comforting phrases such
as: "Learning about equipment is interesting, or: "It's easy."
"

But they did provide a scenario for success: although knowl-

edge on equipment is hard to obtain, it can reduce break-
downs by 23%. This scenario was realized the following year,
and breakdowns caused by operators' insufficient knowledge
dropped from 23 to 4%.
Next, we decided to tackle breakdowns of poorly designed
machines. The previous year's data showed that 17% of total
breakdowns could be directly attributed to poor design. This
figure might be underestimated because the foremen may
have wanted to spare the designers some embarrassment. We
spent seven months redesigning and improving those
machines.
Tofacilitate this process, a system called the "equipment
patrol was devised, it required the designers to patrol the
"

plant and inspect how the machines they had designed were
working. They were expected to spend at least one day per
month following up on the machines' performance, ease of
maintenance, and the level and variability of product quality.
This system was a variation on the QC Patrol which had 1

established in the latter half of the three years when was still
1

manager of the Quality Assurance (QA) section. A QC Patrol

 1

Situational Analysis 1 6

team consisted of a QA manager who acted as leader, and four

to five members from other sections. The team patrolled one
plant each month, inspecting the QC system of the plant from
the customer's viewpoint.
Since the members also had many other responsibilities,
they were allowed to take business trips and leave for meet-
ings during the patrol week, if necessary. The patrol team was
requested to present the result to the manager of the division
as well as managers of the plant and other concerned sections.
The inspected plant had to report to management, within one
month's time, on its improvement plan for any problems that
had been detected.
Redesigning by engineers, in addition to the education of
operators, reduced the total number of breakdowns by 35%,
which was nearly what we had expected. We learned from this
experience that correct situational analysis leads to correct
action.
Another lesson that I learned was that the project man-
ager must present a scenario for success based on clear
reasons and factual data, it is the key to true leadership.
chapter 8

AN INVITATION
TO SELF-STUDY—
FOR THE ACTIVE MIND

8.1 ON THE HAWTHORNE STUDY

In when Drs. Deming and luran first brought QC to
the 1950s,
lapan, we were shocked by differences between Japan and the
United States. Since then, we have been working quite hard to
catch up, and now, thanks to our efforts, and blessed with
favorable conditions, our quality levels have greatly improved.
Strangely enough, our efforts were stimulated by a misun-
derstanding. We thought that the concepts being taught to us
by Drs. Deming and luran were actually practiced by American
firms. Without realizing the differences between research and
practice, and between top-level companies and average com-
panies, we made every effort to catch up with what was, in fact,

an illusion. Theory Y, for example, is not so well known or

applied in the U.S. as in lapan.
Our efforts to catch up brought much success, but we also
realized that hard work alone was not enough to lead us to the
top. Recognizing this, we developed the study group approach
as a means for establishing our own managerial engineering
methods and techniques. Figure 8-1 gives an overview of the
study groups, only some of which are described in this book.
However, hope that this book has communicated the basic
1

163
164 MANAGERIAL ENGINEERING

CREDIBILITY FROM CUSTOMERS

ff
Self-Study 165
 )

166 MANAGERIAL ENGINEERING

ideas and approaches common to all the study groups.

The weakness of blind, hard work is that it may lead to
overlooking the importance of adopting a specific approach.
Professor Lee of Ohio State University criticized the Hawth-
I.

orne Study as a fraud a few years ago. Dr. Marvin E. Mundel, a

former president of the American institute of Industrial En-
gineers, pointed out the failures of the Hawthorne study in a
lecture at Osaka, japan. One of the listeners asked some angry
questions: "When did you discover these Do other re-
errors?
searchers support your theories? Did you know that all
Japanese engineers believe that the Hawthorne study is the
basis for managerial engineering?" For those hard workers, the
Hawthorne study was the symbol of American managerial en-
gineering. What anger they expressed when they learned that
they had been betrayed by a fraud!
The Hawthorne study was discussed in Chapter 3.8. Pro-
is summarized below:
fessor Lee's criticism
(The girls who participated in the experiments were not
1

chosen at random. Among them, two were close friends cho-

sen intentionally
The assembly operation in the test room included far
(2)
fewer parts than the real assembly line, so the work was much
easier.
(3) The girls who were
tested did not perform various ad-
ditional tasks that were assigned to other assembly workers.
(4) The setup operation and supplying of parts were much
facilitated in the test room.

(5) The efficiency rate plan was applied to a group of one

hundred people on the main line, while it applied to five
people in the test room. In addition to better supervision, this
favorable rate plan encouraged the girls to work harder.
(6) On the main line, workers had to go home when a
stock of parts was used up. In the test room, however, stocks
were well maintained.
 Seif-Sludy 167

Lee suggested that it was no wonder that efficiency

Dr.

was much improved under these artificial conditions.

would like to mention two points here. One is that what
I

we learned from the Hawthorne study was not so much the re-
sults of the experiments, but rather, the research attitude that
sought to analyze an unexpected phenomenon and attempted
to discover the unknown factors that caused it. This attitude
was the driving force behind an enormous amount of sub-
sequent research. Even if the Hawthorne study must be
criticized from a methodological point of view, the approach
we learned from it is not weakened as long as we retain con-
fidence in our own self-study. The only point to be corrected is
our understanding of the results of the Hawthorne study.
Another interesting point is that the favorable conditions
that Dr. Lee criticized are just those that have been realized by
Japanese productivity improvement plans. As a result, our pro-
ductivity was improved exactly as Dr. Lee pointed out that the
Hawthorne group's productivity should rise! Professor Lee's
criticism was proven true in Japanese plants.

8.2 IMPROVING THE COMPANY'S CONSTITUTION

lust like a human being, a company has its own constitution. If

a person has constitutional tendencies to catching cold, he

will try to improve his physical condition through daily exer-
cises, taking vitamins, and so forth. Of course, before he can
improve himself, he has to know quite a lot about his physical
constitution. In a company, in order to enhance quality and
productivity, we have to observe ourselves carefully and take
effective actions constantly in our daily work. This may not be
easy, because a constant effort requires a strong will. It may
take much time. However, believe this is a promising way to
I

success. Before concluding this book, would like briefly to in-

I

troduce such an effort which has been made in Meidensha

168 MANAGERIAL ENGINEERING

Electric Manufacturing Co, Ltd.

Through the "Q-up" (quality up) study group activities in
Meidensha, we found that we could use the New loharry's Win-
dow (see Chapter 3.9) as a powerful tool for solving quality
problems. It is now called "Meiden's Window," and is used
company-wide. The philosophy behind it is the same as that of
the OET (On-Error Training) system (see Chapter 6. 1
in which
),

we learn from errors.

Every time we have a defect, whether it is detected within
the company or through a customer's complaint, the informa-
tion is fed back to the responsible section. The section
analyzes the cause of the defect, discusses countermeasures,
and determines what levels of the company's organization will
be involved in preventing the defect from recurring.
Depending on the scope of this projected strategy, we
select appropriate parties for Sections A and B in the Meiden's
Window (Figure 8-2). There are many possible combinations

Figure 8-2 Meiden's Window

 Self-Study 169

for Sections A and B. For example, when the countermeasures

can be taken within one section of the company, the combina-
tion would be section manager vs. members, skilled workers vs. new
workers, etc. When more than one section is involved, we might

choose a combination such as manufacturing department vs. de-

sign department, or Meidensha vs. subcontractors, etc. The important

rule is to select the best combination for facilitating mutual

communication between the two parties and taking the most
effective corrective actions for the defect.
The next step of the responsible section is to record the
defect on the appropriate grid of the Meiden's Window. At the
same time, more detailed information about the defect as well
as the countermeasures are written in the form. When the de-
fect falls into Categories 2 and 3, we use the form shown in Fig-
ure 8-3. When it falls into Category 4, the form shown in Figure
8-4 is used.
At the end of a specified period (usually a month), the re-
sults are calculated and recorded on the table shown in Figure
8-5.

In lune 1982, all the serious company-wide defects and

claims for the past year were classified on Meiden's Window.
The data showed that the ratio among Categories 2, 3 and 4
was about 50%, 25% and 25%, respectively. Since then, we have
been taking the company-wide monthly record, and the ratio
has remained almost the same as in the data above.
Grasping the data led us to an epoch-making change in
our basic strategies for quality problems. The problems in
Categories 2 and 3 require quite different actions from those
we would take for problems in Category 4. With a clear under-
standing of our problem areas, we became capable of imple-
menting the right action for a given problem.
Defects in Categories 2 and 3 can be compared to a com-
mon cold. When a person catches a cold, he often takes some
medicine. He may recover soon, but he is likely to catch
 70 MANAGERIAL ENGINEERING

How to put the knowledge into

practice

Write in how thoroughly we can conduct the coun-

ternneasures each day. Reconsider the contents
- of daily countermeasures if the probability is 70%
Remarks or less.

Figure 8-3 Form for Categories 2 and 3

/'
Within 3 months D
3 to 6 months O
b^
Over 6 months A
Condi- Probability
Number of Counter- Time Required
tions of of
defects measures for Action
defects Achievement

Consider the utilization of the company-wide technical data evaluation system.

Figure 8-4 Form for Category 4

 Self-Study 171

Cate-
gory
POSTSCRIPT

In the present age of slow economic growth, the role of

managerial engineering is bound to continue growing. By
practicing managerial engineering techniques on important
manufacturing problems such as setup time, quality defects
and work-in-process, we learned the following: For managerial
engineering to prove effective in resolving the diverse prob-
lems encountered in manufacturing operations, it must al-
ways look beyond traditional theories through the practical
application of new techniques.
1 have mentioned several actual cases in this book and
stated that such development becomes possible when all the
people who are concerned with a problem participate in its
solution. Only then can their various experiences, knowledge
and creativity be integrated into new methods.
These cases shed light on the methodology we de-
veloped, which is summarized as follows:
• Set an explicit goal and target for the problem
• Extract essential factors necessary for achieving the goal
through practice
• Develop a new theory and method by the integration of

173
174 MANAGERIAL ENGINEERING

these factors with existing knowledge

• Establish a strategy to apply them
• Apply them, analyze the results and make the necessary
adjustments
This book is a collection of cases in which applied the I

above methodology to existing problems, in other words,

managerial engineering is developed through an interdiscipli-
nary systems approach to managerial problems.
have stated that the driving force behind the develop-
I

ment of managerial engineering is self-study. The backbone of

self-study, in turn, is one's firm belief in the knowledge this
pursuit may bring him. "Give me and a lever long
a fulcrum
enough and could move the earth," Archimedes once said. He
1

was not referring to the principle of leverage, but rather, to the

power of the knowledge he had discovered through his
experiments.
In conclusion, I would like to affirm that managerial en-
gineering should seek to improve productivity while unceas-
ingly maintaining the highest ideal of respect for humanity. In
thisage of low economic growth, we must take more care than
ever to observe this highest of goals.
 1

INDEX

Abstract reasoning effectiveness. 39

aid in problem solving, 16-17 example, 34
Accidents explained scientifically, 47
airplane, 20, 22-25 factors affecting applications. 59-6
emergency procedures, 15 implementation. 36
human errors, 18 model of, 57
hydrogen, 15, 20 parameters of. 64
lessons learned from, 13-15 problem-solving ability. 37
prevention, 13, 18 reduction of defects. 44
Airplane accidents research findings, 67-68
human errors, 20 successful groups, 65
lessons learned from, 15. 22-25 use in information gathering, 35
American Society for Quality Control, variations on, 40-42
57 Communication
improving, 4
Bodek, Norman, xxvii model, 47
"Bunt" strategy, 13-114
1 Control chart, 42
Cost
Case study method appraisal, 71
applied to quality control, 148,151 appraisal, budgeting for, 74
CEDAC system failure, 71, 74
analysis of applications, 79-80 potential failure calculation, 74-75
applied to automated plant, 75, 78 prevention, 71, 74
applied to manufacturing process, prevention, budgeting, 75
35-42 quality. 71-75
based on cause-and-effect diagram, 37
conceptual framework, 40-41, 43 Dahlgaard. Dr , 75
conclusions from applications, 67-69 Defects
description, 33 analysis, 143-144
effect of effort, 33 classification, 170-171
effect of participation, 64 human errors, 143
effect of production environment, reduction, 1, 27-28, 44-45, 60, 70
60-61 Deming, Dr W Edwards, 2n, 163
effect of psychological environment, Deming Prize, 2
60-62 Deming Prize Committee, 57

175
 1 3 1

76 MANAGERIAL ENGINEERING

Disaster training standards. 16 motion mind, 121

Donham, Wallace B., 149 origin, 1

science based on humanity, 10

Education setup time reduction, 1 1

favorable work environment, 8 study group approach, 131, 163

industrial engineering course, 120-121 time and motion studies, 1 1

reliability engineering course, 151-152 Industrial Engineering Applications

Equipment maintenance StudyGroup, 84, 91, 102-103
cartoons, 153 Information gathering
importance in production engineer- CEDAC system, 35. 42
ing, 153 Ingram, Harry, 47
improvement in with situational Inventory see stock levels
analysis, 154-159 Ishikawa, Kaoru, 37
Equipment patrol, 160 Itami works, 13, 15

Errors, human
cause of accidents, 17 lapan
cause of airplane accidents, 20 lapan US cultural differences, xxv
cause of defects, 143 lapanese Standards Association, 57
contributing conditions, 18 lapanese Union of Scientists and
preventive measures, 17 Engineers, 149
Ephlin, Donald F xxv , loharrys Window, New
applied to CEDAC, 50
Fail-safe systems explanation, 48-50
guard against human error, 18 formulation, 79
Fei, Han, 68 hypothesis tested, 5 1
-54
Feigenbaum, Armand V used for defect analysis, 143-144
Total Quality Control. 71 luran, Dr , 71, 163
Forecasting
CEDAC model, 57 Kanban system, 84
Kansai Institute of Management and
Harvard Business School, 149 Information Sciences, 91
Hawthorne Study, 46-47, 166-167 Kawakita. liro
Heinrich's principles, 18 How to Get Creative \deas. 78-79
Human errors Kimball, George
cause of accidents, 17 Methods of Operations Research. 107
cause of airplane accidents, 20 Kitamura. Kozo, 20
cause of defects, 143 Kurano, Umakichi. 20
preventive measures, 17
Human Error Study Group, 20 Labor productivity
Hydrogen accidents, 15, 20 increased, 9, 123
Lanchester's Law, 109
IE see Industrial Engineering Lee, Prof I, 166
Industrial Engineering Lighting
and labor productivity growth, 10, 123 relationship to productivity, 46
and robots, 1 1

course, 8, 120-121 Management

importance of practice. 120-121 day-to-day approach, 134, 136
 1 1 1

\ndex 77

determines level of stockless pro- process. 122-123

duction, 103 increased labor productivity, 123
Japanese, xxi-xxii Motivation
policy on production standards, 87 essential conditions for. 21-22
power, 1 means of, 110
reliable methods, 1, 10, 71, 78-79, 107, Multiple regression analysis, 131
136 Mundel, Marvin E., Dr., 166
Managerial engineering OET see on-error-training
adaptation as a tool, 3 On-error-training
as a reliable method, 108 compared to ordinary training, 142,
conditions for implementation, 10 147
difficulty in developing, 68 essential elements, 146
importance of adherence to standards, learning from errors, 141
52-54 similarity toMeidens Window, 168
importance of analytical observations, using errors for practice, 141-146
69 Operations
importance of practice, 10, 108 improvement through motion
need understanding, 9
for clear economy, 126
success through constant effort, 1 72 need to explain observation and
system design, 3 measurement, 124
techniques, 1 Operations Research
through operations research, 83 applied to production control, 82
Manufacturing process managerial engineering technique, 83
CEDAC applied to, 35-42 mathematical models, 83, 94, 109
Mathematical models Methods of Operations Research by Philip
operations research, 83, 94 Morse & George Kimball, 107
production control. 94 origin, 1

Mayo, Elton, 46 OR see Operations Research

Meidens Window
applications, 168-170 Participation
defect reduction, 172 effect on CEDAC applications, 64
on quality control problems, 168
effect should be directed toward clear ob-
Meidensha Electric Manufacturing Co , jectives, 68
xxii.xxiv, 32, 168-169 Practice
Models in emergency procedures, 16
CEDAC system, 57 in management techniques, 10
communication, 47 learning from errors, 14!
mathematical, 83, 94 Problem solving
Morale application of abstract reasoning,
research on factors influencing it, 46 16-17
Morse, Philip application of CEDAC system, 44-45
Methods of Operations Research. 107 application of New loharrys principle.
Motion and Time Study Group. 1 1 52
Motion economy Product
principles, 126-129 failure costs. 75
Motion mind quality. 69
detecting wasteful motions in work quality improvement, 1, 27, 52
 1

178 MANAGERIAL ENGINEERING

Production control through on-error training. 141-146

material flow, 87 through operations research, 10 1

mathematical models, 94-95 through Quality Control Problem

objectives, 81 Study Group. 31
perfect production quality, 70 through stockless production. 85
problems related to stock levels, 91 through waste elimination. 8
standards as management policy. 87 Productivity, Inc , xxii, xxvii

through day-to-day management, 133 Prosperity Through Cooperation

through operations research, 82 Program. 4. 8
through standard flow time, 87 Psychological environment
through stockless production, 102. 106 effect on CEDAC system, 60, 62
through total participation, 103 PTC see Prosperity Through Cooperation
Production engineering
importance of equipment mainte- QC see Quality Control
nance, 153 Quality
Production environment costs, 71-75
effect on CEDAC system, 60-61 Deming Prize, 2
Production Planning and Control Study production, 70
Group, 84, 102 Quality Control
Production system application of case study method to,
control of, 102 148, 15!
kanban, 84 importance of participation, 68
materials flow and operations, 84-85 origin. I

stockless production, 84 popularity in lapan. 27

using standard flow time, 89 problem solving steps. 67-68
using standard production quantity, problems addressed by CEDAC
89 system. 44-45
using standard stock quantity. 89 problems addressed by New loharry
zero defects, 7 principle. 52
Productivity problems caused by variability of
relationship to lighting. 46 procedures. 39
research on factors influencing it. 46 standards. 28
Productivity improvement statistical methods. 35
drum-rim repair operation. 139 through control chart. 42
in operations, 123 through on-error training. 146
labor, 9. 123 Quality Control circles. 33. 44-45
material cutting operation, 138 Quality Control Problem Study Group,
need understanding, 91
for clear 3. 29-32,43. 70-71

of subcontractors, 8 Quality Improvement

stamping operations. 137 products. 27
through CEDAC. 32 through CEDAC system. 32
through day-to-day management, 36 i through on-error training. 141-146
through equipment maintenance. 155 through Quality Control Problem
through group study approach. Study Group. 31
130-131. 133. 135-136
through industrial engineering. 129 Reliable methods
through motion economy, 126-129 importance of practice. 71. 107
 1

Index 79

necessary conditions for. 78 131, 163

scientific methodology for, 78-79 approach to problem solving, 80
search for, 1 for development of managerial
with zero-defects production system, engineering, 3
71 Q-up, 168
Reliability engineering Subcontractors
course final exam, 151 industrial engineering techniques
Robots, 1 1 applied to, 5-7

Ruft, loseph, 47 productivity improvement, 8

Sumitomo Electric Industries
Serling, Robert CEDAC case studies, 32
Loud and Clear, 1 5, 22-25 industrial engineering course, 120
Setup time perfect production quality, 70
reduction, 113, 116 Prosperity Through Cooperation
Shewhart, WA. 42 Program, 4
Shingo, Shigeo, 84, 89 quality control program, 2
Simulations stockless production, 91
production control, 94, 97-99 technical audit system, 158
Situational analysis, 154-160 Suppliers see subcontractors
Standards
conditions for worker adherence, 54 Theory Y, 163
defect reduction, 70 Time and Motion Studies, 123-129
importance of adherence to, 67 Total Preventive Maintenance System,
production control and planning, 87 153-154
quality control, 28 Toyota
standard flow time, 87-89 kanban system, 84
standard production quantity, 89 TPM se? Total Preventive Maintenance
standard stock quantity, 89 System
work, 16 Trainings^? practice
Stock levels
cause of production problems, 91 United States
maximum allowable quantity, 85 US. japan cultural differences, xxv
reasons for accumulation, 93
W-shaped framework, 78-79
Stockless production
Western Electric Co 46 ,

approaches to, 102-103

WIP see Work-in-process
based on principles of production
Work environment
control, 102
favorable to managerial engineering,
development of, 106
4-5,8, 10
just-in-time production system, 84
psychological, 62-63
target level determined by manage-
Work standards, 16
ment, 103
Work-in-process see stock levels
Stress
Workers
as factor in human error, 18
conditions for adherence to standards,
Study groups
54
and productivity improvement, 130,
133, 135-136 Zero defects
approach to Industrial engineering. perfect production quality, 70
About the Author
Ryuji Fukuda's book — a treasure
trove of insightful ideas on industrial
management — offers a new
perspective on the problems that
managers encounter as they try to
boost quality and productivity in their
facilities.
All the methods and techniques
explained in these pages have stood
the test of time. They have been
implemented over the years at
Sumitomo Electric Industries Ltd., one
of lapan's leading industrial groups,
and they have led to outstanding gains
in productivity and quality
improvement.
Ryuji Fukuda was awarded the
prestigious Deming literature prize in RYU|I FUKUDA
1978 for his contributions to the
science of productivity and quality improvement.
Fukuda's career began in 1954 at the R&D division of Sumitomo
Electric industries Ltd. He joined the board of directors at Meidensha
Electric Manufacturing Co Ltd in 1981 and was appointed general
manager of the company's production facilities
Today he is an independent consultant and advisor to Meidensha
Electric and the lapan Management Association as well as to Omark
Industries in Portland, Oregon and other multi-national corporations in
lapan, the US and Western Europe.
Born in 1928, Fukuda graduated from Kyoto University in 1954. He
lectures at Kobe University on reliability engineering.
Fukuda is married and has one daughter.
 " Praise For
MANAGERIAL ENGINEERING
"A great book. Fukuda writes in a very understandable manner and the
translation is the best of any Japanese work that have seen A must' reading
I for
any manager who is at ail interested in improving their operations."
Wayne S. Rieker
""'^
Presidenl. Quality Conlrol Circles. \nc

"Managerial Engineering provides an excellent and clear description of the

devotion and methods of lapanese management to continual improvement of
quality, knowing well that as efforts succeed in improvement of quality,
productivity improves, costs go down With better quality at lower prices, a
company captures the market, provides jobs and more jobs A further important
gain is happier employees —
happy to have the privilege to take part in
company-wide improvement It is a cascading effect Their productivity goes up,
and customers are happier with better quality at lower prices "
W. Edwards Deming
Consultant in Statistical Studies

"A book of exceptional value lam enormously impressed with the

surprising power of simple methods that can be practiced by any manager or
supervisor with a grasp of elementary principles of human behavior and
industrial engineering Fukuda is not an ivory tower academic, but a practical

hands-on operations manager who has achieved wonders in defect elimination,

productivity improvement and enhancement of the quality of work life In
Managerial Engineering Dr Fukuda has been able to communicate his unique
knowledge and skills in a highly readable form so that they can be effectively
used by Americans We are hard at work applying these methods at Omark and
we are delighted with the results so far
lohn L. Warne
President. Omark Industries

"Fukudas book nicely complements the books on just-in-time production

IIT exposes problems and creates the need to solve them right away Actually
solving the problems requires data, and Fukuda shows how to make data
generation for problem-solving an interesting part of the employee's job '

Richard |. Schonberger
and author
Professor. University of biei>raska of
lapanese Manufacturing Techniques

"Ryuji Fukudas book is excellent in all the areas of quality improvement and
IE work The notion of CEDAC deserves wide circulation, and the ideas on
self-learning follow up are great additions to it
'

Robert W. Hall
Associate Professor, \ndiana University and author of
Zero Inventories

Published by:
PRODUCTIVITY,INC.
P.O. Box 16722, Stamford, CT 06905
(203) 322-8388