Simulation-Based Scheduling for a Hot Press Forging Factory Using Data-Driven Models

Abstract

We use a genetic algorithm (GA) to search for an optimal production schedule for a hot press forging factory. Our GA evaluates each candidate schedule by simulating its execution using cost models of all the equipment involved in the forging process. The cost models are learned from data collected from IoT infrastructure installed at our target factory. Experimental results show that our proposed method gives schedules of higher productivity with lower energy cost compared to the heuristic method that is similar to the real practices at our target hot press forging factory.