Ecology Prize of the Viktor & Sigrid Dulger Foundation 2026 goes to Benjamin Schäfer
The Heidelberg Academy of Sciences and Humanities awarded the 2026 Ecology Prize of the Viktor & Sigrid Dulger Foundation to Benjamin Schäfer. The sustainable transition of the energy system is challenging as the availability of renewable power sources like solar and wind energy fluctuates greatly. Benjamin Schäfer is therefore working to increase the stability and resilience of the power grid using transparent, explainable AI methods. He combines physical modeling, machine learning, and open data to quantify fluctuations in the energy system and explain their causes. Together with international researchers, he has developed a data-driven load profile that will make it easier to regulate the balance between supply and demand in the future.
Until now, energy providers have estimated electricity demand using a standard load profile based in part on measurements and consumption data dating back many years. Modern households, on the other hand, have new types of devices such as solar panels and electric cars, which can lead to rapid fluctuations in consumption. “Simultaneously, more operational data is available than ever before. Hence, data-driven approaches become feasible and even necessary,” says Benjamin Schäfer. “Machine learning and artificial intelligence can process these enormous amounts of data, but they must remain explainable. Our algorithm can specify which factors are relevant for its prediction.” The target is to continue operating the power system in a stable and cost-optimal manner in the future. With its Ecology Prize, the Viktor & Sigrid Dulger Foundation supports work that addresses environmental problems and their solutions. The prize has now been awarded to KIT for the third year in a row.
Ecology Prize of the Viktor & Sigrid Dulger Foundation
Data-driven load profiles and the dynamics of residential electricity consumption
published in nature communications 2022
Explainable time-series forecasting with sampling-free SHAP for Transformers
published in nature communications 2026