This paper presents a computational approach to developing design space models that are utilized to improve the design process by predicting values of downstream design attributes based on information available at early stages, such as preliminary design specifications. The predictive models are similar in function, though not in form, to the internal (mental) models created by experienced designers; however, the advantages of these models are that it may be possible to construct them in the absence of a designer’s internal models, and that they can be passed on to and used by less experienced designers. Once created, the computational models aid designers in exploration of design alternatives and to reduce design costs and product development time.

This content is only available via PDF.
You do not currently have access to this content.