Machine intelligence in machine design: Workshop summary

Abstract

This paper summarizes presentations from an invited workshop entitled ‘Machine Intelligence in Machine Design’. This workshop was held on October 3 and 4, 1985 in Ann Arbor, Michigan. The workshop was hosted by the Industrial Technology Institute and General Motors Advanced Engineering Staff. It brought together approximately twenty researchers who have been applying artificial intelligence (AI) techniques to mechanical system design. Participants came from seven universities and two major corporations. A third corporation was represented by a university researcher who had worked on an expert system while on sabbatical at the corporation. The following conclusions were reached at the end of the two-day session: • The mechanical engineering community working in AI should concentrate on the codification of mechanical information for problem description. Mechanical designers should look beyond rule-based expert systems to capture design information; computer scientists should help show the way to proceed. • Meaningful application of Artificial Intelligence techniques to mechanical design cannot be made without standardized data formats for mechanical design elements and assemblies. New search and synthesis techniques for the computer will be required to manipulate the large databases and generate new design concepts and ideas. Computer scientists should concentrate their efforts on the generation of these new AI techniques.