Capturing Design: Improving conceptual ship design through the capture of design rationale

As the complexity of ship design increases, the knowledge used to resolve these complexities is decreasing due to a loss of intellectual resources. In order to remedy these losses, Naval Architects must capitalize on the knowledge available in design teams. The solution to this quandary involves the effective capturing of design rationale. This thesis focuses on the capture of configuration rationale in complex ship design. It recasts the rationale capture problem as a collection of rationale tasks: express, (re)structure, store and (re)use. The methodology developed in this research, integrated in a Rationale Capture Tool, employs three key features. First, a new rationale elicitation approach, Reactive Knowledge Capturing, is used to capture designer reaction to unorthodox and unconventional designs. Second, expressed configuration rationale is (re)structured using a semi-formal argument based ontology that supports indexing, sorting and conversion from a human to machine understandable form. Finally, a dedicated feedback mechanism, which employs an evolutionary search algorithm, triggers designer expression of missing rationales, thereby improving the overall efficiency of the system. A comprehensive test case, with practicing naval architects, illustrates the potential of this approach. It demonstrated the successful capture of configuration rationale (interactions, proximity rules and their associated rationales). A quantitative and qualitative analysis of expressed rationales shows that they are both diverse and appropriate, i.e., relevant and complete. One of the most promising results of the test case was that the majority of interactions and rationales appeared (re)usable, i.e., the rationale can be used to support other types of ship design. The captured rationale serves to improve the quality of knowledge available in early stages of the design process and complement existing conceptual ship design approaches