21 May 2026
Forschungszentrum Jülich (FZJ) is part of the newly established Unit NRW of ELLIS, the European Laboratory for Learning and Intelligent Systems. ELLIS is regarded as one of Europe's leading networks for research in machine learning and artificial intelligence. The new unit brings together leading AI researchers and large-scale computing infrastructure across North Rhine-Westphalia to develop open-source foundation models and datasets and advance trustworthy AI applications for scientific and industrial use.
A central focus of the unit is the development and study of open-source generalist foundation models as core building blocks for machine learning research. It will investigate how such models can be trained to frontier level using supercomputers and open datasets, made more reliable and adapted safely to different application domains. The unit also strengthens research on trustworthy AI, particularly for situations involving uncertainty, rare events or sensitive real-world environments.
A second research track of the ELLIS NRW initiative focuses on transferring machine learning methods into application domains such as healthcare, sustainable agriculture and embodied AI systems like autonomous robots. Researchers in the unit cover fields including robotics, computer vision, natural language processing, healthcare AI and neuro-symbolic systems.
Within the FZJ, the Jülich Supercomputing Centre (JSC), the Institute for Advanced Simulation (IAS) and the Peter Grünberg Institute (PGI) contribute to this initiative. The FZJ is a key contributor alongside Bonn and Aachen in shaping the scientific profile of the Unit NRW, particularly in the area of open foundation models and the datasets required to develop them.
In addition to enabling large-scale training of foundation models, Jülich infrastructures, including JAIF (JUPITER AI Factory), will support the development, deployment and management of large-scale datasets for future open AI systems. This work relies on advanced high-performance infrastructure, including the JUPITER supercomputer at FZJ, and specialised expertise in operating large-scale AI workloads.
"Forschungszentrum Jülich provides a unique combination of AI expertise, high-performance computing and scientific collaboration. Through the ELLIS Unit NRW, we will help advance open foundation model research in Europe and translate AI research into real-world applications."
The initiative will benefit from close collaboration with LAION, an international open research organisation specialised in open foundation models, large-scale AI training workflows and open datasets. LAION has already conducted research on open foundation models using resources from the Jülich Supercomputing Centre (JSC), highlighting the strategic importance of Jülich's high-performance computing infrastructure for open AI research in Europe.
"Through the new ELLIS Unit NRW, LAION will further strengthen its collaboration with ELLIS researchers in the field of open foundation models, building on joint work such as ongoing major European projects including OpenEuroLLM, ELLIOT, Minerva and LLMs4EU", explained Dr. Jenia Jitsev, one of the unit's co-directors as well as scientific lead and co-founder of LAION.
The Unit NRW is coordinated by the University of Bonn under the leadership of ELLIS Fellow Prof. Dr. Jürgen Gall, together with a board of co-directors from the participating institutions, including Dr. Jenia Jitsev (JSC/FZJ & LAION). Additional members from Forschungszentrum Jülich include Prof. Dr. Emre Neftci (PGI), Prof. Dr. Kristel Michielsen (JSC), Prof. Dr. Martin Schultz (JSC) and Prof. Dr. Moritz Helias (IAS).
The unit will include about 40 principal investigators and more than 300 doctoral researchers and postdoctoral scientists across North Rhine-Westphalia, fostering joint research, doctoral training, ELLIS summer schools, as well as collaboration with industry and public-sector partners.
The establishment of the ELLIS Unit NRW further strengthens the role of Forschungszentrum Jülich and the Jülich Supercomputing Centre as key contributors to Europe's open, high-performance and trustworthy AI ecosystem.
