Generalizing the statistical theory of artificial neural networks to biological neural networksJohannes Scmidt-Hieber (Faculty of EEMCS)
Compared to modern AI systems, the brain learns faster, generalizes better to new situations, and consumes much less energy. A child only requires a few examples to learn to discriminate a dog from a cat. And people only need a few hours to learn how to drive a car. AI systems, however, need thousands of training samples for image recognition tasks. And the self-driving car is still under development, despite the availability of data for millions of kilometers of test drives and billions of kilometers of simulated drives.
Why does the brain outperform AI? Artificial neural networks are at the core of the AI revolution. In the past years, enormous efforts have been made to unravel their mathematical properties, leading to fundamental insights and mathematical guarantees on when and why deep learning works well. Artificial neural networks are inspired by the brain but differ in many respects. This project proposes the development of advanced mathematical tools to analyze learning in the brain as a statistical method. If the research is successful, it has the potential to provide insights into how the brain learns and might lead to recommendations on how to make AI more efficient with less data.
ROBOREACTOR: Robotic bioreactors for the longitudinal control of restorative remodelling in the human skeletal muscleProf.dr.ir. Massimo Sartori (Faculty of ET)
Is it possible to regenerate new, healthy biological tissues in the human body after neuro-muscular injuries? With a focus on skeletal muscles and their innervating spinal neurons, the ERC Consolidator Grant ROBOREACTOR addresses stroke-induced neuro-muscular deterioration affecting essential functions like locomotion.
Skeletal muscles exhibit remarkable regenerative capabilities in response to sustained electro-mechanical stimuli. However, muscle regeneration requires a well-orchestrated interaction between the immune system and the musculoskeletal system. After a stroke, disrupted inflammation hinders muscle regeneration, preventing optimal restoration of the lost movement capacity.
Robotic exoskeletons and neuro-stimulators have the potentials to deliver precise electro-mechanical stimuli that could re-optimize inflammatory and remodelling processes in damaged muscles and innervating neurons. The problem is that we don’t understand how robot-delivered stimuli affect muscle inflammation and remodelling across weeks to months.
ROBOREACTOR addresses this problem by creating sensor-driven digital human twins that can predict muscle’s inflammation and remodeling across weeks in response to regimens of electro-mechanical stimuli i.e., tensile strain and compressive stress delivered via soft exosuits and electrical currents delivered via spinal cord stimulation. The idea is to combine AI-powered computer modelling with advanced in vivo neural interfacing and in vitro tissue engineering techniques. The resulting AI-powered digital human twins are finally integrated into predictive control frameworks to create an intelligent neuro-robotic platform that autonomously discovers the stimuli needed for muscles to regenerate optimally over time post-stroke, potentially outperforming conventional manual rehabilitation.
Targeting inflammation and remodeling over weeks, ROBOREACTOR has the potential to reshape stroke recovery, paving the way for intelligent robotic bioreactors and redefining human-robot interaction principles with broad health implications.
WINDFLOW: Unlocking the Complexities of Wind Farm-Atmosphere InteractionDr. Richard Stevens (Faculty of TNW)
Wind farm simulations typically assume ideal atmospheric conditions. However, the dynamics between wind farms and the atmosphere are complex, influenced by factors such as the interaction between the atmosphere and the ocean. Ultimately, unraveling these complex interactions is vital for optimizing wind farm performance. To explore these interactions, Richard Stevens plans to integrate high-fidelity wind farm simulations with large-scale weather models. The goal is to assess and model wind farm performance under real weather conditions, addressing key questions about wind farms’ effects on atmospheric stability, moisture dispersion, and energy exchange between the atmosphere and the ocean. These insights are essential for designing future wind farms.
About the ERCOut of 2,130 candidates, the European Research Council (ERC) has selected 308 researchers for this year’s Consolidator Grants. The funding will support excellent scientists and scholars at the career stage where they may still be consolidating their own independent research teams to pursue their most promising scientific ideas. Worth in total ¤627 million, the grants are part of the EU’s Horizon Europe programme.
The ERC, set up by the European Union in 2007, is the premier European funding organisation for excellent frontier research. It funds creative researchers of any nationality and age, to run projects based across Europe. The ERC offers four core grant schemes: Starting Grants , Consolidator Grants , Advanced Grants and Synergy Grants. With its additional Proof of Concept Grant scheme, the ERC helps grantees to bridge the gap between their pioneering research and early phases of its commercialisation. The ERC is led by an independent governing body, the Scientific Council. Since November 2021, Maria Leptin is the President of the ERC. The overall ERC budget from 2021 to 2027 is more than ¤16 billion, as part of the Horizon Europe programme, under the responsibility of European Commissioner for Innovation, Research, Culture, Education and Youth, Iliana Ivanova.
Student Services Contact Centre
(Internal) Service Portal
Faculties, institutes & service departments