08
Jul
2026
In the first round of funding, projects in the fields of health, climate, robotics, particle physics and sustainable artificial intelligence will receive funding. The aim of the ‘AI Research Groups Lower Saxony’ initiative is to attract and support scientifically outstanding researchers at an early stage in their careers to Lower Saxony, and to expand the development and application of AI in the long term.
The funded AI research groups generally consist of one postdoctoral post for the scientific lead and up to four PhD positions. The projects run for five years, with up to two million euros available per application.
Among the projects receiving funding is Leveraging Artificial Intelligence to Forecast Trends and Extremes in Ocean Temperature, Ocean Productivity and Carbon Cycling at the Carl von Ossietzky University in Oldenburg. Led by Dr Julian Merder, the researchers are investigating how modern AI methods can be used to better predict climate-related changes and extreme events in the oceans. The focus is on marine heatwaves, harmful algal blooms, changes in marine productivity and their impacts on the global carbon cycle. To this end, an AI-supported ‘Digital Ocean Twin’ is to be developed, which utilises global observational data and is capable of forecasting rare and extreme events in particular. The results are intended to contribute to a better understanding of the consequences of climate change and to lay the scientific foundations for sustainable ocean management.
“
The high quality of the proposals submitted has impressively demonstrated just how dynamically AI research is developing in Lower Saxony.
Falko Mohrs, Minister for Science of Lower Saxony
The project Advancing Methods for Infectious Diseases Using Novel Artificial Intelligence Methods at the Helmholtz Centre for Infection Research, led by Dr Alina Bazarova, is developing new AI-based methods to better understand and predict infectious diseases at both a societal and individual level. To this end, health, mobility and behavioural data are analysed using artificial intelligence techniques, deep learning and epidemiological models. The aim is to model the spread of diseases more precisely, identify risk factors and better assess individual risks of disease. The results are intended to improve preparedness for future epidemics, support policy-making and contribute to more personalised healthcare.
Under the leadership of Dr Timo Janßen, researchers at the Georg-August University of Göttingen are developing AI methods for simulating high-energy particle collisions. The aim of their project, Production-Ready AI for Unbiased Simulation in High Energy Physics, is to significantly speed up complex computational processes in particle physics without compromising scientific accuracy. To this end, tried-and-tested AI methods for simulations and model calculations are to be integrated and evaluated within a shared, modular software platform. The results are intended to enable more efficient and sustainable research processes and to advance the use of AI in data-intensive scientific applications.
The project Trustworthy Embodied Foundation Models for Bionic Intelligence at the University Medical Center Göttingen, led by Dr. Sharmita Dey, is developing a new generation of trustworthy AI systems for prostheses and exoskeletons. The goal is to make assistive technologies more intelligent, adaptable, and safer to better support users in their daily lives. To achieve this, a fundamental AI model is being developed that combines visual, physiological, and device-related data, understands situations, and can anticipate actions. Furthermore, the systems should be able to reason over possible actions, recognize uncertainty, and use new evidence to guide decisions.
The project Tuning for FFMs: Optimising Post-Training Pipelines for Energy-Efficient AI at Leibniz University Hannover, led by Dr Marcel Wever, is investigating new methods to make large AI foundation models more energy-efficient. Whilst such models enable powerful automation, their operation is very energy-intensive. The project aims to develop automated procedures that tailor the models specifically to tasks and hardware in order to reduce energy consumption and computational effort without significantly compromising the AI’s performance. The results are intended to help make modern AI applications more sustainable, cost-effective and widely applicable.
A total of 52 applications were received, which were assessed by 16 established AI researchers. For the first time in this call for proposals, applicants had the opportunity to apply for an internationalisation module. This allows applicants to request up to 10 per cent of the actual grant amount additionally to support international collaborations. Two of the funded projects will receive additional support through this internationalisation funding.
A total of 20 million euros is available for this funding programme. In the first funding round, approximately half of the budget will be allocated to five research groups; the remaining funds will be awarded in a further call for proposals in 2027. The funding is provided by the joint funding programme zukunft.niedersachsen.