DigiSWM – AI and advanced data analytics for an interplay between power, heat and mobility

The integration of renewable energy sources in the private sector is made possible in particular by coupling the consumption sectors of electricity, heat and mobility. In order to leverage the enormous potential, energy solutions for the private sector must be consistently thought through. However, sector coupling and the digitization required for it also increase the demands on the actors in the development, parameterization, optimization and marketing of the technologies. Comprehensive energy data (from systems, consumption, and behavior) and AI processes can help to enable new energy services, optimize grid operations, and promote greater dissemination of sustainable energy technologies. The project aims to leverage the potential from existing energy data for such applications. The project's field-tested Big Data Analytics (BDA) toolbox will support households and utilities with machine learning technology to drive sector coupling.

Funded by the Bayerische Verbundforschungsprogramm, Förderlinie Digitalisierung – Informations- und Kommunikationstechnologie

Project duration: 01.07.2021 – 31.06.2024

Research staff at the University of Bamberg: Konstantin Hopf (Principal Investigator), Felix Haag (Project Associate)

Project Partners:

Total project budget: € 1'486'731 (funding and industry contribution), € 218'500 of the total budget for the University of Bamberg

BENEFIZZO – Combining behavioural and analytical innovation to enhance smart meter residential energy savings (machine learning for energy efficiency feedback)

By applying machine learning techniques to large amounts of data on private energy consumption and using advanced feedback technologies, this collaborative project builds a customer engagement platform as a mature prototype. This digital platform enables action-oriented communication tailored to the individual household, which guides private households to greater energy efficiency. Building on the energy efficiency platform to be developed, we will work with energy providers in Europe to create novel services that increase energy efficiency and sustainable energy use in the private sector. In doing so, we address the limits of smart meter technology recognized by companies, research and politics, as well as those of private energy consumption behavior.

The subproject of the University of Bamberg comprises the (further) development and evaluation of machine learning methods that form the basis for the smart energy efficiency platform to be developed. We are advancing machine learning models that can predict household characteristics (apartment size, number of occupants or appliances, etc.) or customers' willingness to participate in energy efficiency or load shifting campaigns using load curves, location information, and other data sources. In addition, we evaluate Explainable AI techniques for their use within the scope of energy feedback.

Funded by the Eurostars Program of the European Union

Project duration: 01.12.2020 – 31.05.2022

Research staff at the University of Bamberg: Konstantin Hopf (Principal Investigator), Felix Haag (Project Associate)

Project Partners:

Total project budget: € 899'680 (funding and industry contribution), € 175'521 of the whole budget for the University of Bamberg

DigiSWM – KI und fortgeschrittene Datenanalysen für ein Zusammenspiel von Strom, Wärme und Mobilität

Die Integration erneuerbarer Energieträger im Privatbereich wird insbesondere durch die Kopplung der Verbrauchssektoren Strom, Wärme und Mobilität ermöglicht. Um die enormen Potentiale zu heben, müssen Energielösungen für den Privatbereich konsequent durchdacht werden. Durch die Sektorkopplung und die dafür erforderliche Digitalisierung steigt aber auch der Anspruch an die Akteure bei Entwicklung, Parametrierung, Optimierung und Vermarktung der Technologien. Umfangreiche Energiedaten (aus Systemen, Verbrauch und Verhalten) und KI-Verfahren können helfen neue Energiedienstleistungen zu ermöglichen, den Netzbetrieb zu optimieren und eine stärkere Verbreitung von Technologien für nachhaltige Energieversorgung zu fördern. Im Rahmen des Projekts soll das Potenzial aus vorhandenen Energiedaten für solche Anwendungen nutzbar gemacht werden. Die im Einsatz getestete Big-Data-Analytics (BDA) Toolbox des Projekts wird Haushalte und Energieversorger mit Machine-Learning-Technologie unterstützen, um die Sektorkopplung voranzutreiben.

Gefördert durch das Bayerische Verbundforschungsprogramm, Förderlinie Digitalisierung – Informations- und Kommunikationstechnologie

Projektlaufzeit: 01.07.2021 – 30.09.2024

Beteiligte Personen an der Universität Bamberg: Konstantin Hopf (Principal Investigator), Felix Haag (Projektmitarbeiter)


Gesamtbudget des Projekts: 1’486’731 € (Förder- und Industriebeitrag), davon 218’500 € für die Universität Bamberg