Reinforcement learning for energy reduction of conveying and handling systems

Schmidl E, Fischer E, Steindl J, Wenk M, Franke J (2020)

Publication Type: Conference contribution

Publication year: 2020


Publisher: Elsevier B.V.

Book Volume: 97

Pages Range: 290-295

Conference Proceedings Title: Procedia CIRP

Event location: Athens IT

DOI: 10.1016/j.procir.2020.05.240


One important reason for wasting energy in manufacturing processes is that individual plant components often consume too much energy despite not being needed for production or nothing being produced at that moment. The implementation of standby-strategies in the PLC to switch each component into an optimal energetic state is applied seldom due to the great programming effort. This is because there are many functional dependencies between the plant components and therefor the PLC program must be adapted manually to the respective manufacturing process. The presented solution shows an intelligent system, which can adapt to plant-specific process requirements and derive decisions about the optimal energetic state of each component autonomously. In this paper, we provide an approach that uses reinforcement learning algorithms, which train on virtual plant models. The results are verified on a conveying and handling system in our learning factory.

Authors with CRIS profile

Additional Organisation(s)

Involved external institutions

How to cite


Schmidl, E., Fischer, E., Steindl, J., Wenk, M., & Franke, J. (2020). Reinforcement learning for energy reduction of conveying and handling systems. In Sotiris Makris (Eds.), Procedia CIRP (pp. 290-295). Athens, IT: Elsevier B.V..


Schmidl, Elisabeth, et al. "Reinforcement learning for energy reduction of conveying and handling systems." Proceedings of the 8th CIRP Conference of Assembly Technology and Systems, CATS 2020, Athens Ed. Sotiris Makris, Elsevier B.V., 2020. 290-295.

BibTeX: Download