My research interest lies enabling robot systems to learn efficiently and combine this with pre-existing knowledge and priors.

The combination of these AI methods with machine-learning techniques such as reinforcement learning is really exciting and enables efficient learning of robust and explainable policies. The current focus lies in contact-rich manufacturing tasks that are solved with manipulators and that require compliant control. My research does not only aim at solving tasks, but also to make this accessible for larger groups, e.g. by easing the problem design and the availability of the algorithms.

I studied Electrical Engineering and Information Technology at the Karlsruhe Institute of Technology (KIT) with a strong focus on robotics, system- and software engineering before starting this position as part of the Wallenberg AI, Autonomous Systems and Software Program (WASP). My involvement in the research program does not only include the research contributions, but also the engagement as a student representative in various committess and building up and learning the core technology cluster "Sequential Decision Making and Reinforcement Learning".

As part of my work I did not only contribute to various public repositories, but also open-sourced key elements of my work such as:

• Cartesian impedance controller: A feature-rich Cartesian impedance controller implementation in C++ that is robot agnostic
  • Paper with the control-theoretical background
• Cartesian trajector generator: Generate and publish Cartesian linear end-effector trajectories with acceleration and velocity profiles
• SkiREIL: "Skill-based reinforcement learning". The algorithms and simulation setup basis for several of the papers.
• ROS Docker containers: Easily usable ROS docker containers with accelerated GUI support