The goal of this research is to determine the technical feasibility of providing expert systems access to external data bases and real time sensory information needed in a manufacturing environment. Most current expert systems interact with people who provide answers to questions posed by the system. For many applications, a user friendly interface supporting textual input/output is adequate; how ever, even the most rudimentary application of expert systems to manufacturing will require more than simple textual input/output capability. Expert systems in manufacturing environments must be capable of accepting real-time sensory inputs from various process variables and using these data as inputs to the inference process. Moreover, such systems will likely require access to remote data bases containing factual information such as inventory data, production needs, and maintenance schedules. Such functions are outside the scope of typical commercial expert system shells, especially those designed for execution on personal computers.
