@techreport{AndraeAuerBarthetal.2020, author = {Andrae, Marc and Auer, Veronika and Barth, Simon and Baumann, Sarah and Betz, Andreas and Ch{\^a}teauvieux-Hellwig, Camille and Geladze, Ekaterine and Hauch, Julia and Hirschm{\"u}ller, Sebastian and Kapitza, Stephanie and Karlinger, Peter and K{\"o}ster, Heinrich and Kramer, Oliver and L{\"u}ftl, Katharina and Martius, Henrike and Mayr, Andreas and Michanickl, Andreas and Miletzky, Frank and M{\"u}hlbauer, Wolfgang and Niedermaier, Peter and Obermeier, Frederik and Peklo, Franziska and Regauer, Verena and Schanda, Ulrich and Schemme, Michael and Schiffer, Lukas and Sch{\"o}pfer, Fabian and Schumacher, Mara and Schutter, Sabina and Schweiger, Andreas and Seckler, Eva and Sigg, Ferdinand}, title = {Jahresbericht 2019, Forschung - Entwicklung - Innovation}, organization = {Technische Hochschule Rosenheim}, url = {http://nbn-resolving.de/urn:nbn:de:bvb:861-opus4-13977}, pages = {60}, year = {2020}, abstract = {Mit dem j{\"a}hrlich erscheinenden Forschungsbericht m{\"o}chte die Technische Hochschule Rosenheim einen Einblick in ihre vielf{\"a}ltigen Projekte und Aktivit{\"a}ten der angewandten Forschung und Entwicklung geben.}, language = {de} } @unpublished{BoehmNeumayerKrameretal.2021, author = {B{\"o}hm, Stefan-Andreas and Neumayer, Martin and Kramer, Oliver and Schiendorfer, Alexander and Knoll, Alois}, title = {Comparing Heuristics, Constraint Optimization, and Reinforcement Learning for an Industrial 2D Packing Problem}, doi = {10.48550/arXiv.2110.14535}, year = {2021}, abstract = {Cutting and Packing problems are occurring in different industries with a direct impact on the revenue of businesses. Generally, the goal in Cutting and Packing is to assign a set of smaller objects to a set of larger objects. To solve Cutting and Packing problems, practitioners can resort to heuristic and exact methodologies. Lately, machine learning is increasingly used for solving such problems. This paper considers a 2D packing problem from the furniture industry, where a set of wooden workpieces must be assigned to different modules of a trolley in the most space-saving way. We present an experimental setup to compare heuristics, constraint optimization, and deep reinforcement learning for the given problem. The used methodologies and their results get collated in terms of their solution quality and runtime. In the given use case a greedy heuristic produces optimal results and outperforms the other approaches in terms of runtime. Constraint optimization also produces optimal results but requires more time to perform. The deep reinforcement learning approach did not always produce optimal or even feasible solutions. While we assume this could be remedied with more training, considering the good results with the heuristic, deep reinforcement learning seems to be a bad fit for the given use case.}, language = {en} }