Startseite » Research » Artificial Intelligence for Energy Systems

Artificial Intelligence for Energy Systems

More powerful data centers, more efficient algorithms and greater data amounts in the increasingly digitalized industry open new possibilities for the use of artificial intelligence. Industrial energy management is complex and tends to become more difficult with increasing energy efficiency requirements. Our vision is that smart software systems assist people with the optimal design, monitoring, operation and maintenance of (energy) systems in production. Therefore, we conduct research in the following areas:

1 | Energy and Market-Adaptive Operational Planning and Optimization

Various influences, highly complex energetic interactions and the integration of dynamic energy markets as well as additional energy costs impede the efficient operation of energy systems through conventional control systems. We survey data and model-based techniques for on- and offline operational planning and optimization. Alongside the „classical“ mathematical optimization, machine learning methods are used.

2 | Forecasts for Energy Demand and Production

A prerequisite for operational optimization and an efficient energy trade is the knowledge about the future energy demand and the self-generation. With different statistical and machine learning methods, causal relationships can be detected and the dynamic of energy flows can be anticipated.

3 | Predictive Component Testing and Maintenance on the Basis of Energy Data

The energy demand of physical systems often correlates with the states of different components. With machine learning, it is possible to detect complex patterns, which might be indicative of component failure or wear. We survey methods with which anomalies on the basis of energy data can be detected and can even be predicted before their occurrence.

4 | Energy Management Expert Systems

In the heterogeneous landscape of production, very individual and complex measures are often identified in the context of energy management. We examine to what extend computer programs can support humans by offering guidance derived from a data base.

5 | Image Recognition to Support Energy Management

Additional data can be created in different application areas with the use of cameras, which can support energy management. We examine the potential of image recognition and information processing for implementation.

6 | Human-Machine-Interface

The interaction with energy data and an energy management system can be very complex and sophisticated. We conduct research approaches in the field of assisting human-machine-interfaces in order to support the energy management.


  • Energy and Market-Adaptive Operational Planning
  • Prognosis for Demand and Generation of Energy
  • Predictive Component Testing and Maintenance on the Basis of Energy Data
  • Energy Management Expert Systems
  • Image Recognition to Support Energy Management
  • Human-Machine-Interface