Conceptual Design: inside and Outside

This paper gives an insight to some of the intrinsic issues related to the understanding and the computer support of conceptual design. On the one hand, the academic research and development have produced innumerable methods and techniques to support product conceptualization. Reasonable progress has been achieved in (a) the understanding of the fundamental reasoning mechanisms, (b) the development of dedicated aspect models, (c) the application of artificial intelligence techniques in computer support, and (d) clarifying the role of conceptualization in the global product development process. On the other hand, the industry still follows intuitive methodologies and applies less-sophisticated techniques to solve conceptualization problems. With a few exceptions, it is rare when a company uses computer aided conceptual design methods and tools from the academia as a daily routine. The author has tried to explore the reasons of this phenomenon. He has become convinced that the problems originate in the endeavor of the academia to introduce abstract models, to develop highly specialized non-integral tools, and to give preference to automated, rather than highly interactive means. He finds the solution in the development of methods, tools and representations that feature very low level of abstraction, but high level of knowledgeintensiveness. The chunks of knowledge of designers have to be taken into account as patterns of conceptualization that describe the design and other concepts as well as the associations among concepts. Any model created in the course of conceptual design should be transferable to CAD/E systems without any loss of data, and without the need for extreme user involvement. Natural communication means (e.g., speaking, gesturing) have to be used to exteriorize the concepts. In handling, manipulation and further elaboration of concepts, we need smart, knowledge-intensive software agents.