Currently, we are experiencing an increasing commercialization of methods that can be roughly summarized under the general term of Artificial Intelligence. The process industry is not ignoring this changes and is investing time and manpower in researching the use of AI. To this end, many project partners have joined together in the KEEN project 04/2020 – 60 percent funded by the German Federal Ministry for Economic Affairs and Energy with a total funding of 23 million euros. INOSIM is a member of several such working groups.

An INOSIM Simulation Environment For Machine Learning

In this context, the KEEN project partners Bayer and INOSIM are working on scheduling production facilities using reinforcement learning. The goal is the very fast generation of production schedules based on self-learned decisions. Our simulation software provides a fast and cost-effective environment to enable such learning. For this purpose, an interface has been developed, with which interventions in the simulation sequence are possible at desired times. The corresponding return of values from the simulation is also possible. This allows the INOSIM simulation to be coupled to custom and standard implementations (e. g., in Python).

Utilizing Artificial Intelligence For Optimized Scheduling

Based on a Bayer case study modeled in INOSIM, we were able to show that a fast generation of schedules is possible. The necessary training for the algorithm can take place before scheduling. The results so far are promising, and further work is underway to improve the scaling of the methodology. With the currently developed INOSIM Core Edition and further work on Reinforcement Learning Algorithms, we will continue to work to ensure that INOSIM is also a good support for our customers´ optimization problems (see also our research project OptiProd).

Would you like to learn more about this project? Are you looking for an innovative SME as research partner? Please contact us.

Array ( [posts_per_page] => 3 [post_type] => [category__in] => Array ( [0] => 66 ) [orderby] => rand [order] => ASC )

Direct Contact

During local business hours

Germany +49 231 97 00 250

USA +1 214 663 3101

India +91 9766 331 092