TY - GEN
T1 - Reinforcement Learning of Dispatching Strategies for Large-Scale Industrial Scheduling
AU - Tassel, Pierre
AU - Kovács, Benjamin
AU - Gebser, Martin
AU - Schekotihin, Konstantin
AU - Kohlenbrein, Wolfgang
AU - Schrott-Kostwein, Philipp
N1 - Funding Information:
This work was partially funded by KWF project 28472, cms electronics GmbH, FunderMax GmbH, Hirsch Armbänder GmbH, incubed IT GmbH, Infineon Technologies Austria AG, Isovolta AG, Kostwein Holding GmbH, and Privatstiftung Kärntner Sparkasse. We are grateful to the anonymous reviewers for their constructive and helpful comments.
Funding Information:
This work was partially funded by KWF project 28472, cms electronics GmbH, FunderMax GmbH, Hirsch Armbänder GmbH, incubed IT GmbH, Infineon Technologies Austria AG, Isovolta AG, Kostwein Holding GmbH, and Privats-tiftung Kärntner Sparkasse. We are grateful to the anonymous reviewers for their constructive and helpful comments.
Publisher Copyright:
© 2022, Association for the Advancement of Artificial Intelligence.
PY - 2022/6/13
Y1 - 2022/6/13
N2 - Scheduling is an important problem for many applications, including manufacturing, transportation, or cloud computing. Unfortunately, most of the scheduling problems occurring in practice are intractable and, therefore, solving large industrial instances is very time-consuming. Heuristic-based dispatching methods can compute schedules in an acceptable time, but the construction of a heuristic providing satisfactory solution quality is a tedious process. This work introduces a method to automatically learn dispatching strategies from just a few training instances using reinforcement learning. Evaluation results obtained on real-world, large-scale instances of a resource-constrained project scheduling problem taken from the literature show that the learned dispatching heuristic generalizes to unseen instances and produces high-quality schedules within seconds. As a result, our approach significantly outperforms state-of-the-art combinatorial optimization techniques in terms of solution quality and computation time.
AB - Scheduling is an important problem for many applications, including manufacturing, transportation, or cloud computing. Unfortunately, most of the scheduling problems occurring in practice are intractable and, therefore, solving large industrial instances is very time-consuming. Heuristic-based dispatching methods can compute schedules in an acceptable time, but the construction of a heuristic providing satisfactory solution quality is a tedious process. This work introduces a method to automatically learn dispatching strategies from just a few training instances using reinforcement learning. Evaluation results obtained on real-world, large-scale instances of a resource-constrained project scheduling problem taken from the literature show that the learned dispatching heuristic generalizes to unseen instances and produces high-quality schedules within seconds. As a result, our approach significantly outperforms state-of-the-art combinatorial optimization techniques in terms of solution quality and computation time.
UR - http://www.scopus.com/inward/record.url?scp=85137896939&partnerID=8YFLogxK
U2 - 10.1609/icaps.v32i1.19852
DO - 10.1609/icaps.v32i1.19852
M3 - Conference paper
AN - SCOPUS:85137896939
T3 - Proceedings International Conference on Automated Planning and Scheduling, ICAPS
SP - 638
EP - 646
BT - Proceedings of the 32nd International Conference on Automated Planning and Scheduling, ICAPS 2022
A2 - Kumar, Akshat
A2 - Thiebaux, Sylvie
A2 - Varakantham, Pradeep
A2 - Yeoh, William
PB - Association for the Advancement of Artificial Intelligence (AAAI)
T2 - 32nd International Conference on Automated Planning and Scheduling
Y2 - 13 June 2022 through 24 June 2022
ER -