5 DAY COURSE
on
TAGUCHI
METHOD FOR QUAL-COST OPTIMIZATION
by
Dr. P. R. Apte
(TIFR, Mumbai)
Challenge :
The liberal economic policy of globalisation has,
for the first time, exposed the Indian industry to global competition as has
never been done before. Large multinationals are purchasing hefty equity stakes
in large Indian companies, revamping and restructuring them to suit their own
product ranges. The liberal economic policy has thus thrown a challenge - to
produce world class quality products at globally competitive prices.
India :
The Indian industry desperately requires Product Design to optimise the thousands of existing products and processes.
The industry could be saved by focusing on Product Design and must improve the
quality of its products and simultaneously reduce their costs drastically
through product (process) design optimisation. It
must do this quickly!
Taguchi Method :
Taguchi Method is a new engineering design optimisation
methodology that improves the quality of existing products and processes and
simultaneously reduces their costs very rapidly, with minimum engineering
resources and development man-hours. The Taguchi Method achieves this by making
the product or process performance "insensitive" to variations in
factors such as materials, manufacturing equipment, workmanship and operating
conditions. Taguchi method makes the product or process robust and therefore is
also called as ROBUST DESIGN.
Taguchi Method is an appropriate solution :
It attacks existing products and processes with a high success rate.
- Competent R&D engineers
can easily apply it.
- Results are achieved quickly
- within 4 to 6 weeks.
- Solutions given by Taguchi
Method can be implemented at an affordable cost.
- It is readily amenable to computerisation.
Engineering Tool and NOT a
Management Tool:
Taguchi method is an experimental engineering tool and hence the success of
its application does not depend on the company environment - cultural changes
are not really required.
Taguchi Method and ISO-9000:
The ISO-9000 aims at improving the capability of an organisation
as a whole to manufacture products to specified technical specification and
quality standards and to deliver them to the customer on time. Taguchi Method,
on the other hand, attacks the product design itself. Through product and
process design optimisation it improves product
quality and reduces costs drastically. Taguchi Method and ISO-9000 thus
complement each other.
Course Objective:
The course objective is to explain the basic philosophy of Taguchi method,
principles of quality engineering, its applications and economic benefits, its
relevance to Indian industry, its successful application in India through
presentation of case studies. Course aims at motivating the industry to learn
to apply this important method for improving the product / process quality and
reap the economic benefits.
Who should attend:
Directors, Vice Presidents, General Managers, Managers, Senior
Engineers concerned with technical matters such as quality, design and
development, manufacturing, production in any industry.
Duration of Course:
This will be a Monday to Friday, 5-Day, 8 hours a day of lectures and
computer software demonstrations.
Course handouts:
There are no course notes as such. One B/W copy of overhead transparencies
will be provided to the organisers (for making
requisite number of copies for distribution to the participants). The text
book, which is an excellent reference for the subject, is
M.S. Phadke
"Quality Engineering using Robust Design"
Prentice Hall (1989)
5 day course
of
Lectures on Taguchi Method
CONTENTS
1.
INTRODUCTION
1.1
The Challenge
1.2 Price of
Ignorance
1.3 Why this
Ignorance?
1.3.1
Research & Development (R&D) and Product Design
1.3.1.1
Research & Development
1.3.1.2
Product Design
1.4
The West, Japan and India
1.4.1
The West
1.4.2 Japan
1.4.3
India
1.5 Indian Situation
1.6 Solution -
Robust Design
1.7 What is ROBUST DESIGN
1.8 History of
ROBUST DESIGN
1.8.1
Made in Japan
1.8.2
Secrets Out After 40 years
1.8.3 Dr. Madhav S. Phadke
1.9
Applications and Benefits of ROBUST DESIGN
1.10 ROBUST DESIGN
and Indian Industry
1.11 ROBUST DESIGN
and ISO-9000
1.12 Threat to
ROBUST DESIGN
1.12.1
Misleading Courses
1.12.2 Technical
Lethargy
1.12.3 Fear of
Reprisals
1.12.4 Technical
Egos
1.12.5 Credibility
1.12.6 Secrecy From Competitors
2.0
PRINCIPLES OF QUALITY ENGINEERING
2.1
Fundamentals of Statistics
2.1.1
Statistical Distribution or Frequency
2.1.2
Mean
2.1.3 Variance
2.1.4 Standard
Deviation
2.2
Causes of variations - Noise Factors
2.3 Effect of Noise
Factors - Economic Losses
2.4 Elements of
Cost
2.4.1
Unit Manufacturing Cost and Quality Loss
2.4.2 Elements of
Costs for Robust Design
2.4.2.1
Operating Cost
2.4.2.2
Manufacturing Cost
2.3.2.3 Research
& Development Cost
2.5
The Fraction Defective Fallacy
2.5.1
Examples illustrating Fraction Defective Fallacy
2.5.1.1
Refrigerators
2.5.1.2 Television
Sets
2.5.1.3 Telephone
Cables
2.5.1.4 Automobiles
2.5.2
Conclusion
3.0
PRINCIPLES OF ROBUST DESIGN
3.1 A New Definition
of Quality
3.2 Target
Performance
3.3 Taguchi's
Quadratic Loss Function
3.4 Average Quality
Loss
3.4.1
Refrigerator Example
3.4.2 Average
Quality Loss - Statistical Correlations
3.4.3 Physical
Interpretation of Average Quality Loss
3.5
Methods of Reducing Variance (variations)
3.6 Benefits of
Reducing Variance with ROBUST DESIGN
3.7 Exploiting Non
Linearity
3.7.1
Control Factors and Their Levels
3.7.2 Noise Factors
- Noise Variances
3.7.3 Nonlinearity
and its Exploitation
3.7.4 Visualisation of Exploitation Nonlinearity
3.8
Stages in a Product Life Cycle
3.9 Classical
Design Approach
3.10 Efficiency of
ROBUST DESIGN
3.11 Taguchi's Two
Step Design Procedure
3.12 Classification
of Parameters - P Diagram
3.13 Classification
of Design Problems
3.13.1
Static Problems
3.13.1.1
Smaller-the-better Type Problem
3.13.1.2
Nominal-the-best Type Problem
3.13.1.3
Larger-the-better Type Problem
3.13.1.4
Signed-target Target Problem
3.13.1.5 Fraction
Defective Type Problem
3.13.1.6 Ordered
Categorical Type Problem
3.13.1.7 Curve or
Vector Response Type Problem
3.13.2 Dynamic
Problems
3.13.2.1
Continuous-Continuous (C-C) Type Problem
3.13.2.2
Continuous-digital (C-D) Type Problem
3.13.2.3
Digital-continuous (D-C) Type Problems
3.13.2.4
Digital-digital (D-D) Type Problems
3.14 Signal-to-Noise
(S/N) Ratio-A New Measure of Quality
3.15 Additive Model
3.16 Matrix
Experiment
3.17 Managing
Economics of Quality
4. STEPS
IN ROBUST DESIGN - A CASE STUDY
4.1
Robust Design in 8 steps :
4.1.1
Planning the experiment : Steps 1 to 5
4.1.2 Performing
the experiment : Step 6
4.1.3 Analyzing and
verifying the results : Steps 7 and 8
4.2 A Case Study : Polysilicon Deposition
Process :
4.2.1
The Polysilicon deposition process in VLSI :
4.2.2 Surface
defects and Thickness uniformity :
4.2.3 Quality
characteristics and objective function :
4.2.4 Control
factors and their levels :
4.3 Matrix experiment :
4.4 Data analysis :
4.5 Predict the improvement :
4.6 Verification Experiment :
4.7 Follow-up Experiments :
4.8 More Case Studies :
4.8.1
Drilling hole in a racing car wheel
4.8.2 Improving
quality of N/C machines
4.8.3 Improving
performance of computer system
4.8.4 Improving
efficiency of telecom networks
5.
MARKETING STRATEGIES WITH ROBUST DESIGN,
TOLERANCE DESIGN AND ISO-9000
5.1.
Apply ROBUST DESIGN and optimise quality-cost.
5.2.
Export with this quality-cost optimisation.
5.3.
Ultimate combination, Robust Design and ISO-9000 certification.
5.4.
Improve quality further through tolerance design after ROBUST DESIGN.
Recommended Text Book:
"Quality Engineering Using Robust Design"
by
Dr. Madhav S. Phadke
Prentice Hall, 1989
Go back to Apte's
web-page
Go back to TAGUCHI Page