A feasibility study of an expert system for a space station and experiments in space
- Paper ID
IAF-86-65
- author
- company
Kamakura Works, Mitsubishi Electric Corporation
- country
Japan
- year
1986
- abstract
Our development philosophy and objectives in researching applications ox artificial intelligence (i.e. an expert system) to a space station are briefly described here. Our final goal is to find optimum points of harmony between men and machinery. The true study of artificial intelligence including expert systems is just beginning, and not only theoretical research but also experimental approaches are necessary. Thus, our first step was to construct an experimental model in order to determine the most effective applications of an expert system to a space station and to achieve optimum harmony between men and machines in a space station. The hardware ox' this prototype consists of a minicomputer for data processing and a special inference machine for the expert system. This allocation of tasks takes into account the development of faster online data processing and faster data analysis for diagnosis, and the hardware itself is expected to be miniaturized and lightened for use in an actual space station. The following became clear based on our study using a prototype. An expert system, a part of an artificial intelligence system, has the potential for performing some of the tasks of an expert engineer, assisting his judgement, and reducing the workload of each onboard engineer, aiming at work saving and automation of the free flyer's operation and control and carrying out space experiments at a space station. This expert system is expected to be very beneficial as a supporting system for engineers in need of quick responses and cool judgements in the event that accidents happen during actual operation or experiments.