DyRo-MCTS: A Robust Monte Carlo Tree Search Approach to Dynamic Job Shop Scheduling

ICLR 2026 Conference Submission8458 Authors

Keywords: Dynamic Job Shop Scheduling, Monte Carlo Tree Search, Combinatorial Optimisation, Genetic Programming, Reinforcement Learning

TL;DR: This paper proposes a robust Monte Carlo Tree Search approach for online planning in dynamic job shop scheduling, where the unpredictability of new job arrivals prevents the use of complete problem information during planning.

Abstract: Dynamic job shop scheduling, a fundamental combinatorial optimisation problem in various industrial sectors, poses substantial challenges for effective scheduling due to frequent disruptions caused by the arrival of new jobs. State-of-the-art machine learning methods have been used to learn scheduling policies that can make prompt, robust decisions in response to dynamic disturbances. However, these offline-learned policies are often imperfect, necessitating the use of planning techniques such as Monte Carlo Tree Search (MCTS) to improve performance at online decision time. The unpredictability of new job arrivals complicates online planning, as decisions based on incomplete problem information are vulnerable to disturbances. To address this issue, we propose the Dynamic Robust MCTS (DyRo-MCTS) approach, which integrates action robustness estimation into MCTS. DyRo-MCTS guides the production environment toward states that not only yield good scheduling outcomes but are also easily adaptable to future job arrivals. Extensive experiments show that DyRo-MCTS significantly improves the performance of offline-learned robust scheduling policies with acceptable online planning time. Moreover, DyRo-MCTS consistently outperforms state-of-the-art MCTS algorithms across various dynamic scheduling scenarios. Further analysis reveals that its ability to make robust scheduling decisions leads to long-term, sustainable performance gains under disturbances.

Primary Area: applications to robotics, autonomy, planning

Submission Number: 8458