Application fields of artificial intelligence in the energy sector - a systematic overview
AI is a promising technology to accelerate the energy transition. Participation shall be enabled by customized products and services, the efficiency shall be increased by a higher degree of automation and a greater utilization of the given assets. Several examples applied in research and the industry can be found, but a structured overview of the application fields of AI in the energy sector, possible development paths and its drivers is missing. By the development of a framework to map the applications and an extraction of application examples from the literature, three superordinate clusters are identified with nine application fields for artificial intelligence in the energy sector. In the cluster "general foundations for decision-making", applications for predictions, operation and asset optimization support the market integration of renewables and allow a higher utilization and better longterm planning of the grid, generation plants and storages. In the cluster "maintenance and security", applications enable an efficient, smooth and reliable operation of the grid and of generation assets by predictive maintenance, by assistance services for technical measures and by security measures. The cluster "distribution and consumer services" comprise of applications where benefits for the consumers are created by enabling a better participation at the energy transition. This is done with the help of tailored predictions and optimization advices, customized products and process automation of retailing and customer services. In general, the most used applications in the energy industry are found in the cluster "general foundations for decision- making", as those are straightforward applications based on advanced data analytics (the so called narrow AI). Applications based on more advanced AI methods (e.g. different forms of input and output data like video, audio or other physical data- the so-called broader AI) can be found in the other two clusters. For a broader application of AI in the energy industry, the most important bottlenecks referring to a missing work force, needed technology improvements and adaptions of the regulatory framework which have to be addressed.