Page 121 - Design for Six Sigma a Roadmap for Product Development

96 Chapter Three

importance. Design structure is the set of interrelationships that char-
acterize the design requirements, design parameters, and process vari-
ables. Depending on the context, different formats to convey the
structure such as block diagrams, process mappings, and functional
trees are used, some more popular than others. While some of the mod-
eling is endorsed in Six Sigma, with the DMAIC approach, like
cause–effect matrices and process mapping, the need is more pro-
nounced in DFSS, in particular the characterization (C) phase. Such
modeling will reveal how the design is coupled in the functional
requirements (FRs). Coupling indicates the lack of independence
between the FRs. Coupling of the FRs is a design vulnerability that
negatively affects controllability and adjustability of the design entity.
In addition, coupling will result in reduced reliability and robustness
of the design entity and will complicate finding satisfactory solutions
that meet customer attributes at release and over time. It will surely
impair the Six Sigma design endeavor to achieve unprecedented cus-
tomer satisfaction capability. Many negative scenarios are produced by
coupling. In a traditional design dilemma, the designer tries to resolve
a detrimental problem on a certain CTS by adjusting some of the
process variables (PVs) without paying attention to the effect of the
adjustment on other FRs delivered by the design entity. This ignorance
or negligence complicates the situation and results in trial-and-error
inertia toward compromised or wrong solutions of the initial problem.
In this situation, the creation of new symptoms in the design entity is
not a remote possibility.
The integration of Six Sigma philosophy with scientific design meth-
ods yields a robust DFSS strategy in both theory and application with
many advantages. For example, the employment of abstraction at high
levels of the design structure facilitates decision making toward
healthy concepts, while the use of mathematical formulation and/or
empirical testing at low levels of the structure facilitates the variabil-
ity reduction and design controllability as the “zigzagging” method of
axiomatic design is used. Axiomatic design provides rules to structure
and select design entities that are robust from a conceptual perspec-
tive when the axioms are obeyed. The optimization (O) and validation
(V) phases of DFSS will be easier to execute when a coupling-free
design is conceived. In coupled concepts, this flexibility is slim.
Unfortunately, axiom obedience is not always feasible, usually for
technological, organizational culture, cost, or other constraints. A design
organization may find itself forced to live with some degree of coupling
in some or all of its designed family, at least in the short term, even
when the technology is capable of resolving coupling due, mainly, to
cost constraints. Therefore, the need to improve the capabilities of a