This session explores the integration of artificial intelligence in decision-making processes across manufacturing systems.

INFORMS 2025 AI Track Talk

Chaired the “Artificial Intelligence for Decision-Making in Manufacturing Systems” session in the AI Track at the 2025 INFORMS Annual Meeting in Atlanta, GA.

Thank you to Dr. Jia Liu, Dr. Anyi Li, Dr. Grace Guo, and Dr. Samar Saleh for their contributions and to everyone who attended and engaged.

Reinforcement Learning (RL) has emerged as a promising approach for real-time decision-making in manufacturing systems under uncertainty. However, the lack of standardized simulation environments has hindered the comparability and reproducibility of RL-based methods developed by different research groups. This project introduces a benchmarking framework designed to address this gap by providing a modular, Python-based simulation environment tailored for manufacturing system control tasks. The framework is compatible with popular RL libraries such as Stable-Baselines3 and supports integration with custom algorithms built in PyTorch or TensorFlow. By offering a shared platform for experimentation, this benchmark aims to accelerate progress in RL research for manufacturing and promote fair, transparent evaluation across studies.