By Genichi Taguchi in Japan in 1950’s 80% of the quality gains of Japan International Acceptance
Nov 14, 2014
By Genichi Taguchi in Japan in 1950’s 80% of the quality gains of Japan International Acceptance
Quality means different things to different people
Following Taguchi, the quality of a product is measured in terms of the total loss to society due to functional variation and harmful effects. The loss would be zero for the ideal quality.
LSL USL
BAD BADGOOD
GOAL POST MENTALITYGOAL POST MENTALITY
0m -
0m +
m
L(y)
m m
Distribution of telephone cable resistance.
Initial distribution
After process improvement and shifting the mean.
Resistance ( Ohms / Kilometre)
m
LSL USL
0m -
0m +
m
L(y)
Average quality loss
Q = k [( m - )2 + 2 ]It consists of two components:
•Shift of process average () from the target value (m)
•Spread of the process (2)
S/N ratios are a log-modified form of Average quality loss function
1. Quality should be designed into the product and not inspected into it.
2. Quality is best achieved by minimizing the deviation from a target. The product should be so designed that it is immune to uncontrollable environmental factors.
3. The cost of quality should be measured as a function of deviation from the standard and the losses should be measured system-wide.
Maximizing S/N ratio is equivalent to reducing variance due to various noise factors and hence improves quality during manufacturing, customer usage and aging and simultaneously reducing cost substantially.
S/N ratio
Identify various causes, known as noise factors, that degrade the product (process) performance variations in raw materials and components,
machinery, workmanship, temperature, humidity, loading, etc.
Eliminate the noise factors one by one
Eliminating noise factors always leads to increased costs
Reduction in profitability or loss of market share in the face of global competition
New method of design optimization for performance, quality, & cost
For existing processes, emphasis is on parameter design
Smallest, affordable development cost
All engineering designs involve setting values of a large number of decision variables.
Common approach is to study one variable at a time or by trial and errorEither long and expensive time span for
completing the design / premature termination of the design process
MATHEMATICAL TOOLOrthogonal arrays to study large number
of decision variables with a small number of experiments
NEW MEASURE OF QUALITYSignal to noise (s/n) ratio to predict the
quality from the customer perspective
1. To establish the best or the optimum condition for a product or a process.
2. To estimate the contribution of individual factors.
3. To estimate the response under the optimum conditions.
RESPONSE
NOISE FACTORS
CONTROL FACTORS
SIGNALFACTORS
M
X
Y
BLOCK DIAGRAM OF A PRODUCT/ PROCESS
Z
A1 A2
Time
Size
Many cos. Big or small, high-tech and low-tech have found the method valuable
High quality at a low competitive price while maintaining profit margin
Upfront improvement of quality by design and process development.
Measurement of quality in terms of deviation from the target (loss function).
Problem solution by team approach and brainstorming.
Consistency in experimental design and analysis.
Reduction of time and cost of experiments.
Design of robustness into product/process.
Reduction of variation without removing its causes.
Reduction of product warranty and service costs by addressing them with the loss function.