Implementation of an expert system for production scheduling

Abstract

This paper describes the experimental implementation of a rule based Expert System for production scheduling in a small manufacturing company. An attempt was made to capture the local knowledge and informal heuristics used by the experienced human production scheduler and to reduce these to a set of rules which would operate on a relational database representing current shop floor status and projected future loading; the objective was for the system to produce schedules similar to those of the human. Some background information is given on the scheduling environment in the company concerned, followed by a general description of the Expert System itself. Some experiences and lessons learned from the experimental implementation are then discussed.