Introduction to Robust Design

and Use of the Taguchi Method

 What is Robust Design
Robust design: a design whose performance is insensitive to variations.

Example: We want to pick x to maximize F

Simply doing a trade study to optimize the value of F

F would lead the designer to pick this point

What if I pick this

point instead?
This means that
values of F as
low as this can
be expected!

2
 What is Robust Design

• The robust design process is frequently formalized through

“six-sigma” approaches (or lean/kaizen approaches)

• Six Sigma is a business improvement methodology

developed at Motorola in 1986 aimed at defect reduction in
manufacturing.

• Numerous aerospace organizations that have implemented

these systems, including:
• Department of Defense
• NASA
• Boeing
• Northrop Grumman

3
Example of Lean Activities at NASA

Progress on Ares “Lean” Activities (cont’d)

• Some example results that are being incorporated into mainline efforts:
– Streamlining boards/panels approval process: reduced from 5 to 2 the
number of board approval steps within Ares
– Design reviews process: 39% reduction in time to conduct design reviews
– Time for risk approval: 66% reduction in the time to evaluate and approve
a candidate risk through the risk management system
– Trade studies: 50% reduction in the number of steps to conduct formal
trade studies - from idea to decision
– Task description sheet (TDS) development for ADAC cycles: from 3% to
80% automation

Less Time on Waste……More Time for

Value Added Work

QPMR_hq20070801ecm
Back to Project Summary Quad Chart

4
 Taguchi Method for Robust Design

• Systemized statistical approach to product and process

improvement developed by Dr. G. Taguchi

• Approach emphasizes moving quality upstream to the

design phase

• Based on the notion that minimizing variation is the primary

means of improving quality

• Special attention is given to designing systems such that

their performance is insensitive to environmental changes

5
The Basic Idea Behind Robust Design

ROBUSTNESS ≡ QUALITY

Reduce
Variability

Increase Reduce
Quality Cost

6
 Any Deviation is Bad: Loss Functions

The traditional view states that there is no In Robust Design, any deviation from the
loss in quality (and therefore value) as target performance is considered a loss in
long as the product performance is within quality  the goal is to minimize this loss.
some tolerance of the target value.

Loss = k(x-xT)2

No
Loss Loss
Loss

xLSL xT xUSL x xLSL xT xUSL x

xT = Target Value xLSL = Lower Specification Limit xUSL = Upper Specification Limit
7
 Overview of Taguchi Parameter Design Method

1. Brainstorming

Design Parameters: Variables under your control

2. Identify Design Parameters
and Noise Factors
Noise Factors: Variables you cannot control or
variables that are too expensive
3. Construct Design of to control
Experiments (DOEs)
Ideally, you would like to investigate all
possible combinations of design parameters
4. Perform Experiments and noise factors and then pick the best
design parameters. Unfortunately, cost and
schedule constraints frequently prevent us
from performing this many test cases – this is
5. Analyze Results where DOEs come in!

8
 Design of Experiments (DOE)

Design of Experiments: An information gathering exercise. DOE is a

structured method for determining the relationship between process inputs
and process outputs.

Here, our objective is to intelligently choose the

information we gather so that we can determine the L9(34) Orthogonal Array
relationship between the inputs and outputs with the Variables
least amount of effort Exp.
Num X1 X2 X3 X4
L4(23) Orthogonal Array 1 1 1 1 1
Variables
Number of Exp. 2 1 2 2 2
Number of X1 X2 X3
Variable Levels Num 3 1 3 3 3
Variables
L4(23) 1
2
1
1
1
2
1
2
4
5
2
2
1
2
2
3
3
1
Number of 3 2 1 2 6 2 3 1 2
Experiments
4 2 2 1 7 3 1 3 2
Num of Experiments must be ≥ system degrees-of-freedom: 8 3 2 1 3
DOF = 1 + (# variables)*(# of levels – 1) 9 3 3 2 1
9
 Inner & Outer Arrays

N3 1 2 2 1
N2 1 2 1 2
N1 1 1 2 2
Design Parameters Noise X1 X2 X3 X4 1 2 3 4
Experiment Num

1 1 1 1 1
2 1 2 2 2 y11 = f {X1(1), X2(1),
X3(1), X4(1),
3 1 3 3 3
Experiment Number

Performance
N1(1), N2(1), N3(1)}
Characteristic 4 2 1 2 3
evaluated at the
specified design 5 2 2 3 1
parameter and
noise factor values 6 2 3 1 2 y52 = f {X1(2), X2(2),
7 3 1 3 2 X3(3), X4(1),
N1(1), N2(2), N3(2)}
8 3 2 1 3
9 3 3 2 1

Inner Array – design parameter matrix

Outer Array – noise factor matrix
10
 Processing the Results (1 of 2)

Noise

Design Parameters Experiment Num

Compute signal-to-noise (S/N) for each row

Experiment Number

1 n 2
Performance

S / N i  10 log   yij 

Characteristic Minimizing performance
evaluated at the
specified design characteristic  n j 1 
parameter and
noise factor values

Maximizing performance 1 n 1 
S / N i  10 log   2 
characteristic  n j 1 y 
 ij 

Inner Array – design parameter matrix

Outer Array – noise factor matrix
11
 Processing the Results (2 of 2)

Isolate the instances of each design parameter at each

level and average the corresponding S/N values.

Design Parameters X1 X2 X3 X4
1 1 1 1 1 S/N1
2 1 2 2 2 S/N2
3 1 3 3 3 S/N3
Signal-to-Noise (S/N)
Experiment Number

4 2 1 2 3 S/N4 X2 is at level 1 in
5 2 2 3 1 S/N5 experiments 1, 4, & 7
6 2 3 1 2 S/N6
7 3 1 3 2 S/N7
8 3 2 1 3 S/N8
9 3 3 2 1 S/N9

S / N1  S / N 4  S / N 7
Avg S / NT1 (1) 
3
12
Visualizing the Results

Plot average S/N for each design parameter

ALWAYS aim to maximize S/N

In this example, these are the best cases.

13
Robust Design Example

Compressed-air cooling system example

Example 12.6 from Engineering Design, 3rd Ed., by G.E. Dieter

(Robust-design_Dieter-chapter.pdf)

14
 Pareto Plots and the 80/20 Rule

20% of the variables in any given system control 80% of the variability in
the dependent variable (in this case, the performance characteristic).

Individual design parameter effects

Cumulative effect
0% 20% 40% 60% 80% 100%

X1
X2
X3
X4
X5
X6
20% of the variables
X7
X8
X9
X10

80% of the variability in

the dependent variable

15
 Limitations of Taguchi Method

• Inner and outer array structure assumes no interaction

between design parameters and noise factors

• Only working towards one attribute

• Assumes continuous functions

More sophisticated DOEs and analysis methods

may be used to deal with many of these issues.

ORI 390R-6: Regression and Analysis of Variance

You can easily spend a whole ORI 390R-10: Statistical Design of Experiments
class on each of these topics ORI 390R-12: Multivariate Statistical Analysis

16
 Conclusions

• Decisions made early in the design process cost very little in

terms of the overall product cost but have a major effect on
the cost of the product

• Quality cannot be built into a product unless it is designed

into it in the beginning

• Robust design methodologies provide a way for the designer

to develop a system that is (relatively) insensitive variations

17