Leveraging Synthetic Data and Machine Learning for Shared Facility Scheduling

This research explores the applicability of machine learning (ML) algorithms in addressing key challenges in manufacturing planning and control (MPC), with a specific focus on capacity requirement planning (CRP) and scheduling. To effectively train ML algorithms, a discrete-event simulation (DES) methodology is employed to construct a system model, generating synthetic data through simulations across diverse scenarios. The proposed framework’s efficacy is empirically evaluated through three distinct case studies, involving sequential, parallel, and shared facility layouts. The sequential and parallel layouts assess overall feasibility and capacity requirement planning, while the shared facility layout investigates scheduling within a more complex flexible manufacturing system. The research findings provide compelling evidence supporting the utilization of synthetic data for training ML models, facilitating efficient resolution of facility scheduling challenges in manufacturing.