Design forecasting: a method for performing DFX analyses in complex product design

Abstract

Design Forecasting is modeling methodology for performing DFX analyses such as Life Cycle Assessment and the Modified Westinghouse Method earlier in the design of automobiles and other complex products. Specifically, Design Forecasting uses probabilistic design methods to supplement DFX analyses in a systematic, yet flexible, manner that reduces data collection ambiguity and effort for the design team. In addition, the methodology uses scenario analyses to evaluate potential design decisions or alternatives. Finally, the methodology allows for rigorous review, improvement, and customization of individual, underlying models. In this dissertation, Design Forecasting is developed, illustrated, and validated using two automotive case studies. The first is a material substitution case study used to evaluate Design for Environment metrics for a Ford C-class sedan. The second case study evaluates Design for Assembly metrics for the washer and wiper systems on the same C-class sedan. The research presented in this dissertation is significant because it defines a previously informal or even nonexistent process within design decision-making, and provides a sound framework for future, scholarly research.