AI research issues in chemical and biochemical process engineering
A.J. Morris, G.A. Montague, M. Aynsley, David Alan Peel · 1990
Rule-based expert type systems have probably run their course of academic interest and development. Richer forms of knowledge representation such as qualitative models, detailed quantitative models, pattern classification based models, order-of-magnitude relationships, etc. can be handled efficiently using modern techniques. These will provide for a real approach towards the emulation of human reasoning. AI methodologies will become ever more important as process engineers are faced with the design and operation of increasingly complex processes. This is particularly the case in process biotechnology, in metabolic pathway synthesis and in molecule and gene design. The promises of artificial intelligence methodologies, towards the solution of engineering problems, will remain just promises unless: all available knowledge forms are used; there is genuine interaction between research and industrial scientists and engineers to provide and make full use of this knowledge; and there is an understanding of the importance of knowledge based systems in science and engineering research and education.