Research

Manufacturing systems can achieve higher efficiency and flexibility through increased autonomy, but this requires solutions that work safely under uncertainty while maintaining precision. Traditional control systems assume perfect models and predictable conditions. Real industrial environments have unknown disturbances, changing parameters, and safety constraints where failures can be catastrophic. Our research addresses these challenges through the combination of physics-informed artificial intelligence, domain knowledge, optimization and contol methods for improving various aspects of manufacturing and industrial processes in general, with the aim to enable flexible automation.