Development of manufacturing control strategies using unsupervised machine learning

This research considers the control of manufacturing systems that support job routing and process sequence flexibility. A machine learning system is presented that uses a simulation model of the target manufacturing system to discover opportunistic control rules. Learning is unsupervised and is driven by a genetic algorithm. The learning method requires very little a priori control knowledge. For this presentation, the decision-making agents are the part types being processed. Part types evolve cooperative strategies for selecting the best route through the manufacturing system based on simulated real-time information that describes the state of the system. Results are presented that demonstrate the effectiveness of the approach.