MINERVA: A Reinforcement Learning-based Technique for Optimal Scheduling and Bottleneck Detection in Distributed Factory Operations1

In manufacturing systems, the term bottleneck refers to a component that limits the entire throughput of a system. A number of approaches have been attempted to find out the bottleneck. However, existing solutions have their own limitations, leaving the bottleneck identification still no trivial task. To address this issue, we study Job Shop Scheduling Problems (JSSP) with realistic extension that jobs are enqueued periodically, and proposes a machine learning based solution to such a problem, named MINERVA. MINERVA first finds the optimal resource scheduling for a target interval, based on a model-free reinforcement learning technique. Then, using a classifier made from an artificial neural network, MINERVA identifies the bottleneck resources for each target interval. We evaluated MINERVA on two representative benchmarks and found that MINERVA is able to detect the system bottleneck with high accuracy of 95.2%, which is almost 25% better than the best among the popular bottleneck identification methods.