Multiattribute Decision Making in Centralized and Decentralized Design

In this project, the theoretical and practical challenges inherent in multiattribute decision making in centralized and decentralized design processes will be addressed. The intellectual advancements include developing a robust approach to probabilistic, multiattribute decision making that uses decision makers' stated preferences in a theoretically sound manner. The plan is to develop provable conditions for determining whether decentralized design processes are stable and convergent or not. Once these conditions are developed, the plan is to use them to study how to effectively alter a decentralized decision process in order to achieve better collective solutions than those found by traditional sequential or iterative processes. Since decision making is recognized as a fundamental construct in engineering design, the principles developed in this work will impact a number of scientific communities, including design theory, operations research, design automation, and control theory. In addition, the advancements are important to multiattribute decision scenarios in industrial design processes, from selecting a design concept, to determining a solution configuration, to choosing materials, to selecting suppliers. The results will be applicable across a wide range of design education activities, including sophomore and senior product design courses, graduate level design courses, and workshops on decision-based design. It is also anticipated that the developments will foster further collaborations with researchers in decision sciences, product marketing, and microeconomics. Lastly, the advancements will involve undergraduate research students, high school students through a gifted math program and summer workshops, and minority students in Science, Engineering and Mathematics through two federally sponsored programs.