Adaptive systems will be increasingly used in the future to assist people in the solving of complex tasks and problems. In addition, they help companies avoid errors and use resources in a particularly efficient way. The necessary condition for adaptive systems is the targeted use of sensor data, which is usually collected by the systems in large quantities.
We develop solutions that allow these volumes of diverse data to be processed and interpreted in a way that is particularly efficient and features a large element of self-learning. This leads to optimum strategies for the control of systems.
Our approach is based on the integration of expert knowledge, multi-physical modeling and state-of-the-art methods of data-based system analysis and machine learning.
The results of our work are used in a broad range of industrial fields − from quality management for industrial processes, to learning systems for condition monitoring and predictive maintenance, to the measurement-data-based management of wireless networks, to the optimization of energy systems in buildings and production facilities.