Chatting with wind turbines: How a GenAI chatbot supports data analysis at EnBW

The wind turbines of the energy company EnBW not only generate electricity in daily operation, but also a mass of data. Interactive graphics and analyses have always been available in Asset RADAR, but had to be implemented specifically for each application. GenAI now makes this operating data interactive and flexible: EnBW employees can use a chatbot to create ad hoc graphics and analyses in no time at all. This supports the condition diagnostics of all the company's wind turbines.

EnBW is one of Germany's largest energy companies. Part of its business is the construction and operation of wind turbines. To avoid power generation outages, EnBW's diagnostic engineers continuously monitor the technical condition of the turbines. To do this, they analyse so-called SCADA data, which provides information on speed, power or wind speed. This data diagnosis helps to detect damage in good time and thus to save costs in the maintenance and operation of the wind turbines.

EnBW Asset RADAR, a self-developed condition monitoring system, is used for the diagnosis. Although the application can be used to create analyses and standard graphics, each use case had to be defined and developed in advance – a time-consuming and resource-intensive task. This limited the number of use cases.

Together with adesso, EnBW expanded the Asset RADAR to include a chatbot that enables interaction with the operating data of the wind turbines. Based on Microsoft Azure OpenAI with Model GPT-4o, the chatbot now provides answers to any questions about the operation of the plant. From retrieving individual statistics to calculations and specific analyses, everything is possible. For example, the average temperature of identical wind turbines at different locations in June can be displayed in seconds using a prompt.

The chat with the wind turbines is made even more valuable by the unlimited variety of visualisations. Whether bar charts, heat maps or interactive charts: diagnosticians can create graphics ad hoc without having to develop a specific function in the software or make a request to the business intelligence team.

With the new chatbot, based on Microsoft / Azure OpenAI technology in Asset RADAR, EnBW is providing diagnosticians with a flexible toolbox. Graphics can be created and specific data analyses can be carried out in seconds. The application thus takes the complexity out of the mass of data and provides directly usable insights for diagnostics. This creates a broader knowledge base for better decisions. Without GenAI, it would not have been economically feasible to enable such a flexible output of analyses and visualisations.

What makes the lives of the diagnostic technicians at EnBW's wind turbines easier today can also benefit others in the future. For example, other departments can obtain information from the data of the turbines using the same scheme. Using interfaces (APIs), the use case can be easily transferred – in the energy sector, but also in other industries.