The University of Edinburgh -
Division of Informatics
Forrest Hill & 80 South Bridge

Research Paper #538

Title:Towards Ai-Based Design
Authors:Smithers,T; Tang,M
Date:Jul 1991
Presented:Revised version of a paper accepted for the International Young Computer Scientists Conference "Towrards the Future", ICYCS'91, held in Beijing, The People's Re
Abstract:It is widely recognised that the use of Artificial Intelligence techniques forms the foundation of the next generation of Computer Aided Design systems and Computer Aided Engineering systems. An AI in Design research programme has been conducted in the Department of Artificial Intelligence, University of Edinburgh since 1984. From 1985 to 1990, an AI-based design support system, known as Edinburgh Designer System (EDS), was developed as part of the Alvey large scale demonstrator project "Design to Product". The Edinburgh designer System was built on an exploration-based design model, in which design is considered as a kind of intelligent behaviour. The EDS architecture was implemented as an integrated applications of a number of Artificial Intelligence techniques in a belief that design processes can be intelligently supported by computer systems which effectively combine human expertise and Artificial Intelligence. The EDS was primarily built to support mechanical engineering design and it's architecture and the underlying design model are now being applied to other domains such as molecular design and electric design. In this paper we discuss the prospects of AI-based design research by reviewing the achievements of the Edinburgh Designer System and the problems we have encountered n developing it. The Edinburgh design model is firstly introduced and an overview of the system architecture is then presented. Each component in the architecture is described in view of their roles in the architecture and their functionality in relation to the design model. Finally the research problems arising from present work in the AI in design research programme are discussed. These include knowledge representation, inference control, the maintenance of the knowledge base and the application of AI-based design support system in the design of water turbines.

[Search These Pages] [DAI Home Page] [Comment]