Tradespace Exploration for Decision Making in Conceptual Design of Complex Systems

Abstract

Industries have adopted the practice of using Tradespace Exploration (TE) to investigate cost and performance of various designs for a product. Due to added complexity new tools are required beyond conventional System Engineering methods. Organizations require methodologies to quickly explore many concepts and easily determine those designs which will provide the highest value or utility to the stakeholders of the system. TE is a technique which is used during Concept Exploration phase, which is the first phase of the System Development Life Cycle, to weigh the utility of potential design combinations against the attributes which are deemed important by the system stakeholders. This enables the comparison of multiple potential architectures for a product and a more thorough exploration of the design space. TE leads to cost effectiveness and simplifies rework in Engineering Development phase. This research aims at studying prior models which have been developed for carrying out TE and taking inspiration from them develop a cross-disciplinary framework which explores potential design options for two different case studies: Automated Insulin Pump and a Micro Aerial Vehicle. The stakeholder requirements were translated to a Quality Function Deployment (QFD) to derive significant requirements. A system breakdown hierarchy was formulated to identify the components, the variants for which were enumerated using a Morphological box. A Design Structure Matrix was used to highlight the relationship between design components. The major concerns of this thesis were to formulate the utility functions, generate the design space, transform it into tradespace and discover Pareto optimal architectures. A survey questionnaire was distributed amongst type 1 diabetes patients to find the customer preference for various design components of an Insulin Pump. These preferences are unit normalized so that their value lies between 0 and 1. ASSESS, an interactive tool for conducting stakeholder interview was used to find utilities for Micro Aerial Vehicle (MAV). Architecture enumeration lead to the generation of a design space which was translated into a tradespace. Non-dominated architectures were identified in order to plot the Pareto frontier, which was plotted using regression analysis. The architectures which lie on the Pareto frontier are optimum architectures and these need to be explored further by the decision makers.