Implementation of Machine Learning Algorithms in Advanced Planning and Scheduling (APS) Tools for Setup Time Optimization: A Systematic Review

The optimization of setup times in manufacturing is important to increasing production efficiency and flexibility, especially in highly dynamic environments. Traditional Advanced Planning and Scheduling (APS) systems often rely on static optimization techniques, limiting their effectiveness in handling complex and variable setup processes. This systematic review explores the integration of Machine Learning (ML) algorithms into APS tools to enhance setup time optimization. By analyzing recent studies, we identify key ML approaches, their impact on production scheduling, and the main challenges faced during implementation. The findings highlight that ML-driven APS solutions can achieve setup time reductions of up to 35%, particularly in discrete manufacturing industries. However, challenges related to data quality, system integration, and computational complexity must be addressed to fully leverage these technologies. This review provides a structured framework for the application of ML in APS, bridging the gap between theory and industrial practice, and outlining future research directions for enhancing setup time optimization in manufacturing.