| This dissertation presents an approach toward automating the collaborative design process using the advantages offered by the simultaneous engineering activity. The research focuses on the design and manufacturing of automotive sheet metal parts, and on the selection of steel as workpiece material in particular.;An expert system is developed based on the use of multiple knowledge sources. To provide a powerful knowledge integrating mechanism, a blackboard architecture is adopted as the framework of the multi-expert system. The domain knowledge is organized into design, material, the manufacturing knowledge sources. These knowledge sources communicate through a common working memory to provide partial solutions to the design problem.;A frame-based, object-oriented design representation scheme is devised to construct and manipulate models of sheet metal parts on computer. The models are designed with special primitives called laminas, transitions, and features. It is shown that these primitives form a powerful repertoire well-suited to most sheet metal part designs. Based on the information embedded in the representation scheme, analysis and synthesis methods can be applied to improve the existing design.;The collaborative design process in the multi-expert system is arranged as a hierarchical agenda. Tasks and subtasks that constitute an agenda are assigned priorities. Control strategies are developed to search for feasible and optimal solutions. The automated design process is intelligently controlled since the search strategies are able to dynamically apply various knowledge bases.;The multi-expert system has been implemented by using a frame-based, object-oriented tool running on a high-end workstation. The system, once fully implemented, can be used to train novices, or to assist experienced engineers in sheet metal part design. |
