Page 589 - Design for Six Sigma a Roadmap for Product Development

546 Chapter Fifteen

a systematic approach to achieve this goal. This approach is the robust
parameter design for dynamic characteristics, and it will be discussed
in the next section.

15.3 Parameter Design for
Dynamic Characteristics
15.3.1 Signal-response system
and ideal function
Dynamic characteristic is also called signal-response design. Figure 15.7
gives a block diagram of a signal-response design.
Dr. Taguchi’s use of this signal-response system for quality improve-
ments started with communication and measurement systems, both of
which can be represented by the signal-response system well. For a
measurement system, the signal is the measurement sample; the out-
put response is the measurement value of a sample characteristic. For
a good measurement system, the following conditions are required:

1. From the same sample, the measurements must be repeatable no
matter who is doing the measuring and how many times the sam-
ple is measured.
2. From two different samples, a small difference in true characteris-
tics should be detected by the measurement system; that is, it is
desirable to have high sensitivity.
3. A measurement system must be easy to calibrate; a linear relation-
ship between the true value of the characteristic and measurement
value is ideal.

The ideal relationship between signal and output response (e.g., a
functional requirement) is illustrated in Fig. 15.8.
In a later application of the Taguchi method, the signal-response
system serves as the paradigm for many technical systems. In most
cases, the ideal signal-response relationship is the ideal function that
we discussed in the last section. The robust parameter design for
dynamic characteristics actually becomes the robust parameter design
for ideal functional quality.

Signal System Output
response

Noise
Figure 15.7 A P-diagram with input signal.