Two more ERC Proof of Concept grants for Leibniz University Hannover (LUH): Prof. Dr. Sören Auer and Prof. Dr. Marius Lindauer from the L3S Research Center have received funding to translate the results of their respective ERC Grant projects into real-world applications. Both projects focus on artificial intelligence research.
Proof of Concept (PoC) is a funding programme from the European Research Council (ERC) which can be awarded in addition to the main funding lines for innovative cutting-edge research (Starting, Consolidator, Advanced and Synergy grants). It is intended exclusively for researchers who have already received an ERC grant – one of the most prestigious research funding awards in Europe. PoC grants are intended to help explore the commercial or social potential of a completed ERC project. The funding totals 150,000 euros.
Twenty researchers at LUH are currently conducting research funded through an ERC grant. Ten of them have an ERC Starting Grant, eight have an ERC Consolidator Grant and two have an ERC Advanced Grant. Prof. Dr. Sören Auer completed his ERC Consolidator Grant–funded research in April 2024. Prof. Dr. Marius Lindauer will conduct research with his ERC Starting Grant until November 2027.
AI assistant to help make research faster and more precise
Sören Auer is a professor of data science and digital libraries at LUH and director of the TIB – Leibniz Information Centre for Science and Technology and University Library. With his ERC Consolidator Grant, he researched the AI-assisted analysis of academic publications together with the TIB as part of the ScienceGraph project. His Proof of Concept project aims to show how this research can be implemented as a usable AI research assistant. The assistant is intended to provide answers based on verified research data, always citing the sources, assessing the significance of studies and methods, and clearly indicating uncertainties. The project will be tested using two specific applications: One is located within the field of materials science. Here the assistant will provide guidance on test reports and the comparison of hypotheses. The other is located within the field of data science. Here the project will focus on reproducibility – that is, whether methods, data sets and analyses are easily comprehensible and reproducible. “Within a period of 12 months, the assistant should reach a level of maturity that allows it to be used at universities, libraries and research institutions and in industrial research and development,” said Auer. “The goal is a trustworthy AI research assistant which performs research more quickly without compromising academic precision.”
AI development focused on people
Prof. Dr. Marius Lindauer is a professor of machine learning at the LUH Institute of Artificial Intelligence. His ERC Starting Grant focuses on the continued development of automated machine learning (AutoML). The complete harnessing of AI’s potential depends on how quickly and reliably new AI applications can be developed. AutoML can support the efficient development of these applications. However, today’s AutoML systems usually function like rigid “black boxes”. This means developers cannot easily contribute their expertise, development takes longer, new ideas emerge more slowly and ambitious AI solutions to complex tasks are difficult to implement. The Proof of Concept project AutoML Co-Designer is developing an initial, functional prototype for a new, joint form of AI development. The goal is to transform AutoML from a purely automatic tool to an interactive system in which AutoML and developers work closely together. “If this concept is successful, it will form the basis for a new type of AI development tool,” said Lindauer. He explained: “The Proof of Concept shows how ideas can be translated more quickly into functional applications. It opens up new markets for specialised AI solutions that have previously been too complicated or expensive to develop. In doing so, the project lays the groundwork for a market-ready product that helps development teams gain an edge in the rapidly growing AI economy.”
