Optimizing schedules for adaptable manufacturing systems

Abstract

Adaptability and changeability are becoming the key features of future manufacturing. With adaptability comes a need to dynamically and frequently change factory topologies throughout the life-cycle of manufacturing systems. These changes are necessary to respond to changes in markets that result in new products, production volumes, and product variants. In addition to flexibly switching between different setups, the resulting schedules still have to be optimal to a certain extent to stay profitable. In this paper, we extend a capability-based planning approach to generate timed automata models that are used to solve the scheduling problem. The capability-based approach automatically generates an action sequence that defines all necessary production resources and operations for a product. The action sequence is then automatically transformed into models in UPPAAL - a model-checker for timed automata - and used to generate a schedule for production orders. The result can then be executed on the different production resources. An example scenario based on an industrial production system used for educational purposes shows the suitability of this approach.