A decision-friendly approximation technique for scheduling multipurpose batch manufacturing systems

This paper presents a method for computing sub-optimal schedules for certain complex manufacturing systems. These systems are used to produce various products using a fixed factory infrastructure. Such systems can be relatively easy to model and simulate, but computing optimal schedules for them is often intractable. Here, we sample a simulation of the complex system to develop an integer program approximating the complex relationship between admissible schedules and their overall performance, measured in terms of system throughput. Solving this integer program yields a sub-optimal schedule that seems close to optimal on examples