Dynamic Lot Size Optimization with Reinforcement Learning

Research output: Contributions to collected editions/worksChapterResearchpeer-review

Authors

Production planning and control has a great influence on the economic efficiency and logistical performance of a company. In this context, this article gives an insight into the use of simulation as a virtual model of a filling machine in the process industry. Furthermore, it shows the possibilities of a reinforcement learning (RL) approach for dynamic lot sizing. The contribution indicates a possible implementation in an ERP system and shows how a decision support tool can support the planner to save up to 5% of costs compared to a human planner and a heuristic approach proposed by Groff.

Original languageEnglish
Title of host publicationDynamics in Logistics : Proceedings of the 8th International Conference LDIC 2022, Bremen, Germany
EditorsMichael Freitag, Aseem Kinra, Hebert Kotzab, Nicole Megow
Number of pages10
Place of PublicationCham
PublisherSpringer Science and Business Media B.V.
Publication date2022
Pages376-385
ISBN (Print)978-3-031-05358-0
ISBN (Electronic)978-3-031-05359-7
DOIs
Publication statusPublished - 2022

    Research areas

  • Lot sizing, Reinforcement learning, Simulation
  • Engineering