Representation Methods for an Axiomatic Design Process Software

Four types of knowledge representation, relational, inferential, inheritable, and procedural knowledge, are investigated within Axiomatic Design (AD). These knowledge types provide the framework to analyze existing representation methods and propose additional methods to be automated by the Axiomatic Design Software (ADS) currently being developed by the AD Group at MIT. Case studies are presented to illustrate the AD method and indicate the need for structured representation methods for design knowledge within industrial design projects. The ADS is being designed within the axiomatic design process. The functionality and program elements of the software, including the database and graphical user interface, are presented. The software automates the proposed representation schemes to document the development of new design information, based on given design constraints and higher level decisions. Additional research on creating databases for each design domain (customer, functional, physical, and process) is required to develop a thinking design machine (TDM) which can recommend solutions. The structure of each supporting database is dependent on the representation scheme of the AD process. Thus, the representation scheme presented in this thesis can contribute to the development of a TDM. Thesis Supervisor: Nam P. Suh Title: Ralph E &. Eloise F Cross Professor & Mechanical Engineering Department Head