Page 214 - Design for Six Sigma a Roadmap for Product Development

DFSS Transfer Function and Scorecards 187

design structure. For example, the functional requirement (FR) of a
given subsystem or component can be the input signal of another com-
ponent delivering another FR, and so on. These relationships create the
design hierarchy. Uncoupled hierarchy is desirable to accomplish the
high-level FRs of the design and can be examined using a functional
block diagram. The block diagram is a graphical representation of the
design mappings and noise factors effects. An example is depicted in
Fig. 6.1. In process/service-oriented Six Sigma projects, the block dia-
gram is actually a process map. This section provides the foundation for
diagramming by linking the physical structure components using
mathematical formulation and Taguchi robustness concepts (e.g., noise
versus design factors) and tools (e.g., P-diagram; see Sec. 5.9.1).

6.3.1 P-Diagram
The robustness approach is based on a revolution in quality improve-
ment because of the application of certain statistical methods to opti-
mization by Dr. G. Taguchi. The principle idea is that statistical testing
of products should be carried out at the design stage, in order to make
the product robust against variations in the manufacturing and usage
environments.
Using this methodology, quality is measured by statistical vari-
ability, such as standard deviation or mean-square error rather than
percentage of defects or other traditional tolerance-based criteria.
The objective is to keep performance on target value while minimiz-
ing the variability. Robustness means that a design performs its func-
tion as intended under all operating conditions throughout its
intended life. Noise factors are the undesirable and uncontrollable
factors that cause the FRs to deviate from target values. Noise fac-
tors adversely affect quality. However, it is generally impossible or
too expensive to eliminate noise factors. Instead, through robust
design, the effect of noise factors can be reduced.
Robust design is aimed at reducing the losses due to variation of
performance from the target value based on the quality loss function,
signal-to-noise (S/N) ratio, optimization, and experimentation. The
design output is usually categorized into a desired part containing a
useful portion of FR and extraneous or undesired portion. In the
dynamic case, the desired portion is called “signal” and the undesired
segment is called “error.” Usually, both are added together to form the
total output achieved. The primary objective of robust design is to
reduce the effect of noise as much as possible.
The P-diagram (Fig. 6.2) of robust design represents all the ele-
ments of synthesis activity of the DFSS algorithm. The components of
the P-diagram are: