Performance vs. cost analysis: A structured methodology for quantitative design concept selection

Selection of a design concept is one of the most critical, but the most difficult decision in product development due to the large degrees of uncertainty in the preliminary design phase. Many methodologies attempt to quantitatively identify the most preferred concept; however, the majority are deterministic methodologies that do not analyze uncertainty. Decision Analysis is a probabilistic methodology that model uncertainties relevant to the decision. In product development, Decision Analysis has a significant contribution when setting target requirements and cost of a product that serve as the basis to evaluate design concepts. To integrate Decision Analysis and concept evaluation, the existing quantitative methodologies need modifications.; The author proposes a structured methodology called Performance vs. Cost Analysis (PCA) that integrates the following three steps: (1) Identify customer needs and constraints; (2) Set target requirements and cost using Decision Analysis; (3) Identify the most preferred concept based on target requirements and cost, and the importance of customer needs.; To integrate these steps, the PCA modifies existing tools, and presents new applications using the next generation linear collider currently developed by Stanford Linear Accelerator Center (SLAC) as an illustrative example. In the grouping of customer needs, the PCA compares the groupings by Affinity Diagram and by Subjective Clustering to test if any subjective bias exists in the consensus-based grouping. Decision Analysis identifies the target requirements and the budget that optimize the overall objective of the SLAC that is the realization of the linear collider and its potential for the scientific discovery. For the concept evaluation and selection, first, the PCA decomposes Quality Function Deployment (QFD) for the components of the linear collider and their requirements, and deduces the importance of these components and requirements from the importance of customer needs. Then, the PCA modifies Pugh's rating/weighting method to evaluate design concepts two-dimensionally by their performance relative to the target requirements, and the degrees of satisfying the target cost. This two-dimensional analysis prevents one to accidentally choose a concept with very high performance but unacceptably high cost, or a concept with very low cost but unattractive performance for customers.