7. June 2022

Data Science Joins Process Engineering At The 5th KEEN Project Meeting

Frankfurt/Main (Germany) was the place of the 5th project meeting of KEEN, which was held on May, 20, 2022. Within KEEN, INOSIM, in cooperation with partner Merck, is developing a batch production support system that harnesses modern AI methods and the established INOSIM Simulation Software Suite to provide adaptive real-time predictions that will make the complex behavior of modern batch plants transparent to the plant operators and production managers. In another cooperation, the partners Bayer and INOSIM are developing KI-based methods for production scheduling.

High-End Research Results

Twelve PhD students presented their current results. From AI-based image analysis to the standardization of metadata, all topics of the subprojects were presented. INOSIM employee Dominik Bleidorn gave a presentation on the automatic calculation of reactive production plans for industrial batch plants using the AI method Reinforcement Learning. The presentation included an introduction to the application of Deep Reinforcement Learning in reactive production planning and highlighted opportunities and challenges. Preliminary results showed that production planning can benefit from the fast response times of AI-assisted methods.

Connecting Industry And Academia

The programme was supplemented by lectures from industry and academia. In practical terms, the focus was on the use of digitization and data for industrial transformation, using Air Liquide as an example. More theoretically, current research in the field of Deep Anomaly Detection was another lecture´s subject. The breaks in beetween were used extensively for the exchanges between the participants.

Nicolas Weiner, the technical advisor of the DLR Project Management Agency, concluded by praising the outstanding results and the successful event. Prof. Dr. Leon Urbas, coordinator of KEEN, closed the event with the motto: “The process industry gives meaning to the data scientist.”

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

INOSIM Support

During usual business hours

Germany +49 231 97 00 250

USA +1 214 663 3101

support@inosim.com