Interview with Prof. Dr. Michael Bortz on the second funding phase of the »AI-FOR« research group / 2026
AI Alone Isn’t Enough – It’s the Combination of Models, Simulation, and Process Knowledge That Makes It Powerful
How can errors and critical conditions in chemical plants be detected early on – even when only limited data is available? This is the question being addressed by the research group »AI-FOR: Deep Learning on Sparse Chemical Process Data«, funded by the German Research Foundation (DFG). Following successful results in the first funding phase, the DFG has now approved an additional five years of funding for a second phase. The research group has already developed new deep learning methods for anomaly detection and, for the first time, made extensive open datasets for chemical process data available. In addition to the University of Kaiserslautern-Landau (RPTU), the Technical Universities of Munich and Dortmund, as well as our institute, are participating in the project.
more info