Reference: Engelmore, R. S. & Tenenbaum, J. M. Toward the Engineer's Associate: A National Computational Infrastructure for Engineering. KSL, August, 1990.
Abstract: This paper, which was originally written as a "White Paper" for DARPA, proposes a new initiative to develop concepts, techniques, tools and knowledge bases that will increase the sophistication and efficiency of engineering practice in the United States, and reduce life cycle costs of engineered artifacts. Specifically it is argued that within the next 3 to 5 years the nation should develop a computational infrastructure for enginnering practice and education, through a combination of: (1) large, sharable knowledge bases containing concepts, empirical principles of engineering, process knowledge, and catalogs of designs and components; (2) tools for leveraging engineers' skills in design, analysis, diagnosis and explanation, including powerful prototyping environments, case- based and model-based reasoning methods, and integrated qualitative and quantitative simulation; and (3) innovative computer systems that help engineers and managers work together in both large and small groups ("collaboration technology"). In the long-term (10 to 30 years), this research and development effort will lead to the creation of artifacts capable of solving a wide range of engineering problems with minimal human assistance, thereby alleviating the critical shortage of skilled practitioners in all engineering disciplines. Accomplishment of both near-term and long-term goals would constitute a revolution in engineering practice.
Full paper available as hqx, ps.