In the latest call for the prestigious ERC Advanced Grants, the University of Stuttgart celebrates a remarkable success: two researchers have been awarded this highly competitive European research funding. Among them is Prof. Dr. Blazej Grabowski, Participating Researcher in the Cluster of Excellence SimTech and head of the project network “Data-Integrated Design of Functional Matter Across Scales,” whose project META-LEARN received a €2.5 million grant from the European Research Council.
Grabowski, Professor at the Institute for Materials Science and head of the Materials Design group, is internationally recognized for his work on quantum-based simulations of materials. With META-LEARN – Meta-Learned Machine-Learning Interatomic Potentials for Ab initio Engineering of Chemical and Microstructural Complexity, he aims to unlock a new era in material design by using artificial intelligence to navigate the complexity of atomic interactions.
Cutting the Gordian Knot of Material Simulation
At the heart of META-LEARN lies an ambitious vision: to make the construction of machine-learning interatomic potentials (MLIPs) faster, smarter, and more accessible. These potentials – mathematical models describing atomic interactions with quantum-level accuracy – have shown tremendous promise in recent years. Yet, their practical use has remained limited due to the high complexity involved in designing, training, and tuning them for real-world materials.
Grabowski’s project will change that. By developing a meta-learning framework that systematically captures expert knowledge from a wide range of simulations, META-LEARN will generate a comprehensive AI-driven system – the MLIP Co-Pilot. This tool will recommend optimal combinations of algorithms, parameters, and training data for the design of accurate and efficient MLIPs across diverse materials challenges.
“MLIPs have the potential to revolutionize material science,” says Grabowski. “But only if we manage to make their development scalable and user-friendly. That’s exactly what META-LEARN is about.”
Scientific Depth with Real-World Relevance
Beyond algorithmic innovation, META-LEARN is deeply connected to practical challenges in energy and sustainability. The project will apply its methods to complex materials systems such as hydrogen storage alloys and protective coatings – areas where chemical and microstructural intricacies demand a new level of simulation power.
By integrating domain expertise from quantum mechanics, thermodynamics, and machine learning, the project also aligns with SimTech’s core mission: combining data-driven methods with high-fidelity simulations to design functional matter across scales.
Third ERC Grant Underscores Scientific Excellence
Blazej Grabowski is no stranger to the ERC. Following an ERC Starting Grant in 2015 and a Consolidator Grant in 2019, this third ERC award confirms his leading role at the interface of materials science and simulation technology.
Alongside Grabowski, physicist Prof. Laura Na Liu was also awarded an ERC Advanced Grant for her project in DNA nanotechnology.
Prof. Peter Middendorf, Rector of the University of Stuttgart, congratulated both recipients of the ERC Advanced Grants: “Professors Laura Na Liu and Blazej Grabowski are setting international standards in their fields. These grants demonstrate the excellence and global visibility of our strategic research areas – Biomedical Systems, Quantum Technologies, and Simulation Science.”