A scheduling algorithm for dynamic distributed manufacturing maintenance

This paper presents a scheduling algorithm for resources (maintenance engineers) in distributed dynamic manufacturing maintenance environments. The algorithm is designed as a distributed search procedure and implemented in multiagent systems architecture. As in today's manufacturing maintenance environments the overlapping structure of engineers' skill sets provides a great degree of flexibility in dealing with dynamic changes, we formally model the flexibility-degree in our scheduling approach and use it to direct the optimization behavior of our algorithm. In the design of the multi-agent distributed scheduling system, we focus on two major aspects: (1) the negotiation protocol for which we apply contract net iteratively to assign operations to engineers and (2) the agents' local decision making using heuristic local scheduling algorithms based on composite dispatching rules. The effectiveness of the approach Is demonstrated through a computational study.