Page 129 - Design for Six Sigma a Roadmap for Product Development

104 Chapter Three

Morphological approaches to synthesis developed by Zwicky (1948)
and Hubka and Eder (1984) are very similar to the different effects
and analogies presented in TIPS. In these approaches, a complex
design problem can be divided into a finite number of subproblems.
Each solution of a subproblem can be considered separately. Solutions
are then arranged in charts and tables. The morphological charts and
matrices have been developed to suggest possible solutions or avail-
able effects that can be used in a certain situation. Most of the charts
and matrices are developed for mechanical product design, and may be
difficult to use outside their intended fields. Hubka and Eder (1984)
and Ramachandran et al. (1992) researched the synthesis problem and
focused on automating the synthesis process. To automate synthesis,
most researchers have limited their applications to a certain field. In
doing so, only a few principles are covered. Many automated approaches
have been implemented as tools using artificial intelligence (AI). They
are, however, specific to one or a few engineering principles. It appears
to be difficult to find solutions based on other principles, using
these tools, such as the building block approach of Kota (1994). The
approach to analyze a solution in most product development research
is based on comparison. Matrices are commonly used to represent the
engineering situation. The matrices can be arranged in different ways:
the comparative criteria on one axis and the solution on the other,
functional requirements on one axis and the proposed solution on the
other axis, or solution decomposition on both axes. The comparative
approaches of Clausing (1994), Pugh (1991), Ullman (1992), and Phal
and Beitz (1988) are most commonly used. These matrices can be used
in situations where solutions to be evaluated originate from the same
principles and the same objectives.
In the axiomatic design approach suggested by Suh (1990), evalua-
tion can be made by analyzing how well-proposed solutions are fulfill-
ing the functional requirements. This approach enables evaluation of
solutions based on different principles. The main advantage of evaluat-
ing matrices with selected solutions on both axes is the possibility of
sequencing or scheduling design activities. In this area much research
has been conducted by McCord and Eppinger (1993), and Pimmler and
Eppinger (1994). Algorithms for optimizing and resequencing project
structure are some of the results of this research category. The strength
of this evaluation technique is in the sequencing and optimization of
engineering projects. In these situations only limited support is offered
by sequencing methods to the synthesis of new solutions.
In the United States, and since the early 1970s, there have been pro-
gressive research efforts in the design arena, particularly in the field
of mechanical design. Engineering design research was motivated by
the shrinking market share of the United States. The engineering