Jannik Grothusen

I'm a graduate student at TU Hamburg, where I study Mechatronics with a focus on learning-based control. I've spent two semesters at UC Berkeley and now I'm doing research on robot learning at TUM.

In 2023, I was part of Entrepreneur First's W23 Grad cohort in London, where I worked on real-world applications for robotic learning.

Before that, I was building robotic kitchens at GoodBytz as one of their first employees. Also, I worked on autonomous electron accelerators at DESY.

Email  /  CV  /  Twitter  /  LinkedIn  /  GitHub

Research

I'm interested in computer vision, machine learning, and control theory. Most of my research is about learning-based control for physical systems.

A data collection framework for robot learning of manipulation tasks.

Supervisors: Daniel A. Dücker & Sami Haddadin - MIRMI @ TU Munich

Finetuning Octo to a new observation/action space for ALOHA/ACT tasks.

Integration of attention mechanisms into MPC for a learnable control policy.

Supervisors: Tony Zheng & Francesco Borrelli - MPC Lab @ UC Berkeley

Reinforcement learning for the automatic start-up of electron linear accelerators.

Supervisors: Oliver Stein & Annika Eichler - MSK @ DESY

A control architecture for autonomous grasping with a micro underwater robot.

Supervisors: Daniel A. Dücker & Robert Seifried - MuM @ TU Hamburg

Miscellanea

Presenting my thoughts on Google DeepMind's Robotic Transformer RT-2 alongside its lead author Yevgen Chebotar to a group of interested students.

Co-organizing a conference to foster dialogue among individuals from European and US politics, government, businesses, think tanks, academia, and students.

This website is based on Jon Barron's source code.