Ph.D. Proposal Presentation by
Matthew K. Chamberlain
Tuesday, July 12, 2005
(Dr. Farrokh Mistree, Chair)
"An Approach to Decision Support for Strategic Redesign"
As a natural result of invention, changing markets, competition, and other forces, products and whole product portfolios tend to evolve with time. By some estimates, as much as 90% of all design activity involves some sort of adaptive or variant design. Most research in design theory, on the other hand, is concerned with supporting “clean sheet of paper” projects. In complex systems, a lack of decision support for redesign means that systems are oftentimes redesigned in a quasi-haphazard way, guided at best by expert experience and intuition. The danger in this approach is that seemingly simple redesigns in one subsystem may have system-wide effects the designer could not have envisioned, sometimes even making future redesign targets harder to achieve.
The focus of this work is on the creation and demonstration of a systematic decision-based method to support the identification of high-level parameters in the process of redesigning an existing system to meet a series of goals representative of emerging market niches. The approach taken is to consolidate what would be an otherwise independent series of redesign decisions hierarchically using a modified form of the Product Platform Constructal Theory Method (PPCTM), which has been implemented previously for designing families of customizable systems. While the basic tenets of constructal theory remain valid, the PPCTM has been modified so that family synthesis decisions are based on three new criteria relevant to a redesign process: target achievement, similarity to the existing product, and commonality across the whole family of systems. To provide insight into which redesign options merit exploration in PPCTM and how these should be organized hierarchically, a new step is added. In this step, design of experiments (DOE), response surface methodology (RSM), and design capability indices (DCI) for measuring robustness are combined to create a method for quickly exploring and visualizing the potential impact of redesign options. This method would fill a gap in design theory in providing the ability to support designers in making redesign decisions strategically with respect to both current and future goals.