We are looking for a dexterous manipulation expert to strengthen our team in Zurich. The goal of this position is the development and deployment of state of the art learning-based controllers to interact with the environment and manipulate objects more reliably and flexibly.

Requirements

As all code has to run on real robotic hardware in real time, hands-on robotic experience and knowledge in state of the art models for dexterous manipulation, such as either Diffusion Models, Flow Matching or Reinforcement learning is required. 

Experience with ROS and perception models is a plus.

Particularly:

• PhD or master's degree in robotics with relevant project experience (e.g. master thesis) on data-driven manipulation or diffusion models.
• Excellent knowledge of Python, PyTorch and the training of deep neural networks.
• Hands-on experience for running learning-based controllers on real robot hardware.