A LEARNING REACTIVE SCHEDULER USING CBR L

Machine learning techniques can increase the power/expertise and impro ve the problem-solving efficiency of artificial intelligence applicati ons for complex problems such as reactive scheduling in manufacturing operations management. Reactivity requires adaptively changing behavio ur of a performance system that can best be supported by learning from experience. Reactive scheduling is considered in this paper with two options for reactive repair/proactive schedule adjustment to compensat e for/prevent the effects of disturbances during the execution time of a current schedule. With a supervisory control perspective to shop fl oor operations, reactive scheduling of the two options above is embodi ed as one major intelligent supervisory function in a supervisory reac tive scheduler SUPREACT and complemented by a cognitive learning appro ach to improve search space/control pattern-matching knowledge for gui ding iterative schedule repair actions. In a brief overview, some lear ning approaches that comprise past experience for future re-use in rea ctive scheduling are referred to, The paper then presents a combined r ule-and case-based reasoning/learning approach to opportunistic reacti ve scheduling based on a blackboard framework of the system's Expert S upervisor Unit, which also supports human integration into supervisory control of executed processes. Inherent learning ability of the case- based component allows to capture new schedule repair/search control k nowledge, including also human preferences, and thus improve the syste m's reactive/proactive schedule repair efficiency in response to unexp ected events/performance deterioration trends during the execution of predictive schedules in manufacturing shop floors. (C) 1997 Elsevier S cience B.V.