
M. Sc. Felix Thömmes
- Research Associate
- Group:
- Room: 107
- Phone: +49 721 608-43236
- Fax: +49 721 608-42707
- felix thoemmes ∂does-not-exist.kit edu
Karlsruher Institut für Technologie (KIT) Campus Süd
Institut für Regelungs- und Steuerungssysteme
Geb. 11.20 (Engler-Villa)
Kaiserstr. 12
D-76131 Karlsruhe
Curriculum Vitae
Dual study of electrical engineering (Bachelor) at Trier University of Applied Sciences in cooperation with Westnetz GmbH, Bachelor thesis 2022 on the topic of collision detection of mobile robots. Subsequently study of electrical engineering and information technology (Master) with a specialization in systems engineering at the Karlsruhe Institute of Technology (KIT), Master thesis 2024 on trajectory reconstruction with spiking neural networks. Practical work at the Research Center for Information Technology (FZI) in the field of simulation and modeling of AI accelerators. Since October 2024 research assistant at the IRS.
Research
Non-Equilibrium Game Theory for Cooperative Systems
Robots that work directly with humans are on the rise in many areas, for example in medical technology, care and industrial production. To ensure that this collaboration is safe and efficient, humans must be specifically taken into account when designing and controlling such systems.
As human movements can often be described as solutions to optimisation problems, it makes sense to also optimally adapt the robot to this behaviour; an approach that is being investigated in so-called dynamic game theory. However, before a state of mutual optimality (also known as equilibrium) can be achieved, humans and machines must first get to know each other and adapt to each other. What makes it exciting is that both sides adapt simultaneously: a dynamic process of mutual co-adaptation.
In my research, I am concerned with the systematic description of such learning processes. I am investigating how human learning behaviour can be modelled, whether and under what conditions the learning processes converge to an equilibrium and how the robot can be designed in such a way that it specifically supports this process. The aim is to make human-robot collaboration safer, more efficient and more intuitive.
If you are interested in topics such as human-robot collaboration, optimisation or adaptive control for your thesis, I look forward to hearing from you!
Title | Type |
---|---|
Non-Equilibrium Behavior in Dynamic Games through Gradient Play | Master Thesis |
A Game-Theoretic Learning Model Connecting Nash and Stackelberg Equilibria | Master Thesis |
Experimentelle Untersuchung des menschlichen Lernprozesses in Dynamic Games | Bachelor/-Master Thesis |