{"citation":{"bibtex":"@article{Fischer_Hüsener_Grumbach_Vollenkemper_Müller_Reusch_2024, title={Demystifying Reinforcement Learning in Production Scheduling via Explainable AI}, DOI={<a href=\"https://doi.org/10.48550/ARXIV.2408.09841\">10.48550/ARXIV.2408.09841</a>}, journal={arXiv}, publisher={Cornell University}, author={Fischer, Daniel and Hüsener, Hannah M. and Grumbach, Felix and Vollenkemper, Lukas and Müller, Arthur and Reusch, Pascal}, year={2024} }","mla":"Fischer, Daniel, et al. “Demystifying Reinforcement Learning in Production Scheduling via Explainable AI.” <i>ArXiv</i>, Cornell University, 2024, doi:<a href=\"https://doi.org/10.48550/ARXIV.2408.09841\">10.48550/ARXIV.2408.09841</a>.","short":"D. Fischer, H.M. Hüsener, F. Grumbach, L. Vollenkemper, A. Müller, P. Reusch, ArXiv (2024).","alphadin":"<span style=\"font-variant:small-caps;\">Fischer, Daniel</span> ; <span style=\"font-variant:small-caps;\">Hüsener, Hannah M.</span> ; <span style=\"font-variant:small-caps;\">Grumbach, Felix</span> ; <span style=\"font-variant:small-caps;\">Vollenkemper, Lukas</span> ; <span style=\"font-variant:small-caps;\">Müller, Arthur</span> ; <span style=\"font-variant:small-caps;\">Reusch, Pascal</span>: Demystifying Reinforcement Learning in Production Scheduling via Explainable AI. In: <i>arXiv</i>, Cornell University (2024)","chicago":"Fischer, Daniel, Hannah M. Hüsener, Felix Grumbach, Lukas Vollenkemper, Arthur Müller, and Pascal Reusch. “Demystifying Reinforcement Learning in Production Scheduling via Explainable AI.” <i>ArXiv</i>, 2024. <a href=\"https://doi.org/10.48550/ARXIV.2408.09841\">https://doi.org/10.48550/ARXIV.2408.09841</a>.","ama":"Fischer D, Hüsener HM, Grumbach F, Vollenkemper L, Müller A, Reusch P. Demystifying Reinforcement Learning in Production Scheduling via Explainable AI. <i>arXiv</i>. 2024. doi:<a href=\"https://doi.org/10.48550/ARXIV.2408.09841\">10.48550/ARXIV.2408.09841</a>","apa":"Fischer, D., Hüsener, H. M., Grumbach, F., Vollenkemper, L., Müller, A., &#38; Reusch, P. (2024). Demystifying Reinforcement Learning in Production Scheduling via Explainable AI. <i>ArXiv</i>. <a href=\"https://doi.org/10.48550/ARXIV.2408.09841\">https://doi.org/10.48550/ARXIV.2408.09841</a>","ieee":"D. Fischer, H. M. Hüsener, F. Grumbach, L. Vollenkemper, A. Müller, and P. Reusch, “Demystifying Reinforcement Learning in Production Scheduling via Explainable AI,” <i>arXiv</i>, 2024."},"title":"Demystifying Reinforcement Learning in Production Scheduling via Explainable AI","year":"2024","type":"journal_article","publication_status":"published","status":"public","_id":"5470","main_file_link":[{"url":"https://arxiv.org/abs/2408.09841","open_access":"1"}],"language":[{"iso":"eng"}],"abstract":[{"lang":"eng","text":"Deep Reinforcement Learning (DRL) is a frequently employed technique to solve scheduling problems. Although DRL agents ace at delivering viable results in short computing times, their reasoning remains opaque. We conduct a case study where we systematically apply two explainable AI (xAI) frameworks, namely SHAP (DeepSHAP) and Captum (Input x Gradient), to describe the reasoning behind scheduling decisions of a specialized DRL agent in a flow production. We find that methods in the xAI literature lack falsifiability and consistent terminology, do not adequately consider domain-knowledge, the target audience or real-world scenarios, and typically provide simple input-output explanations rather than causal interpretations. To resolve this issue, we introduce a hypotheses-based workflow. This approach enables us to inspect whether explanations align with domain knowledge and match the reward hypotheses of the agent. We furthermore tackle the challenge of communicating these insights to third parties by tailoring hypotheses to the target audience, which can serve as interpretations of the agent's behavior after verification. Our proposed workflow emphasizes the repeated verification of explanations and may be applicable to various DRL-based scheduling use cases."}],"user_id":"220548","publication":"arXiv","date_created":"2025-01-30T13:40:38Z","date_updated":"2025-01-31T12:27:46Z","author":[{"last_name":"Fischer","full_name":"Fischer, Daniel","first_name":"Daniel"},{"first_name":"Hannah M.","last_name":"Hüsener","full_name":"Hüsener, Hannah M."},{"orcid":"0000-0001-6348-7897","orcid_put_code_url":"https://api.orcid.org/v2.0/0000-0001-6348-7897/work/177055139","last_name":"Grumbach","full_name":"Grumbach, Felix","id":"243801","first_name":"Felix"},{"id":"245570","first_name":"Lukas","full_name":"Vollenkemper, Lukas","last_name":"Vollenkemper"},{"full_name":"Müller, Arthur","last_name":"Müller","first_name":"Arthur"},{"first_name":"Pascal","id":"223271","full_name":"Reusch, Pascal","last_name":"Reusch"}],"oa":"1","publisher":"Cornell University","doi":"10.48550/ARXIV.2408.09841"}