User-Centered AI Methods for Energy Applications

The Competence Center for “User-Centered AI Methods for Energy Infrastructures” focuses on the development and application of artificial intelligence methods for the planning, operation, and transformation of integrated energy infrastructures. The focus is on municipal, industrial, and infrastructure stakeholders who are confronted with increasing complexity, growing data availability, and new demands on decision-making processes in the context of the energy transition.

A key focus of our work is the further development and integration of AI into existing modeling and planning tools for energy infrastructures. In particular, we are further developing the open-source software pandapipes and expanding it into pandapipesAI to directly link physical network models with artificial intelligence methods. 

Building on this, we develop and integrate agent-based AI systems (Agentic AI) that automatically combine data analysis, simulation, and evaluation, and are incorporated into planning processes as active assistance systems. These new forms of interaction are complemented by approaches based on explainable artificial intelligence (Explainable AI) to ensure transparency, traceability, and trust in AI-supported decisions. The developed approaches are applied in various use cases, including the integration of energy infrastructures with deep geothermal energy, cross-sectoral heat planning, the development of resilient energy infrastructures, and cybersecurity issues in coupled electricity and heating networks.

Another key focus is the integration and analysis of heterogeneous data sets. Using the concept of the municipal data funnel, we are developing approaches to systematically consolidate and process different data sources and make them usable for planning and decision-making processes. The goal is to bring together data, models, and knowledge in integrated, user-centered systems and make them accessible to various stakeholders.

Our work aims to bring together complex models, data, and knowledge in integrated, user-centered systems and make them accessible to various stakeholders. In doing so, we place particular emphasis on transparency, traceability, and supporting well-informed decisions.