First Database on Solar Tower Power Plants Published

June 18, 2026

Source: E & M powernews

KIT and DLR are making operational data from a solar tower plant publicly available for the first time. AI applications and digital twins could drive further developments in the technology.

Solar tower power plants could benefit from new AI applications and digital twins in the future. Researchers at the Karlsruhe Institute of Technology (KIT) and the German Aerospace Center (DLR) have published, for the first time, a freely accessible dataset containing operational data from the Jülich solar tower plant.

Unlike photovoltaic systems, solar tower power plants do not convert sunlight directly into electricity. Instead, movable mirrors—known as heliostats—focus the sun’s rays onto a receiver at the top of a tower. The heat generated there can be stored, used to generate electricity, or utilized for industrial processes. By storing the heat, solar tower power plants can supply energy even outside of daylight hours.

From the “PAINT” database—whose content is readable by both humans and machines—researchers can download data for individual heliostats or specific time periods and integrate it directly into machine learning models. The data can also be used to develop digital twins of solar tower power plants that virtually replicate real-world facilities.

“Such digital twins make it possible to first test power plant operations using a simulation model,” says Daniel Maldonado Quinto of the DLR: “If we combine them with machine learning, we’ll be able to determine in real time whether the mirrors are correctly aligned and how the power plant’s control variables need to be adjusted for safe and efficient operation.”

849 gigabytes of data

According to the authors, “PAINT” comprises 849 gigabytes of operational data from the Jülich solar tower plant covering the years 2021 through 2024. This includes information on 2,014 heliostats, their positions and movements, as well as more than 218,000 images for analyzing radiation focusing. The data is supplemented by measurements of mirror geometry and weather data.

A key focus of the research is the precise alignment of the heliostats. Even minor deviations caused by wind, material wear, or control errors can reduce the plant’s performance or place a strain on components. With the help of AI models and digital twins, such effects are to be better detected in the future and control strategies optimized.

“Operating solar tower plants safely and efficiently is complex and expensive,” says Kaleb Phipps of KIT’s Scientific Computing Center. “To develop and reliably test new methods, researchers need real-world operational data. Our database makes such data available in an open and structured format.”

The database emerged from work on the “ARTIST” research project, an AI-supported model for developing digital twins for solar tower power plants. Researchers from KIT, DLR, and the Helmholtz AI platform were involved, and the results were published in the journal *Nature Energy*.

Author: Katia Meyer-Tien