Robustify Behaviour and Morphology of Robots against Future Damage

Scope: Master’s Project/Master’s thesis

Advisor: Dr. Kevin S. Luck (

Added: 31.August 2022

Topic: Recent work in deep reinforcement learning sought to co-design the behaviour and morphology of robots in a joint fashion. AI technology developed in this area will help to support human designers and engineers in their quest to find new robot designs. This project is aimed at developing new methodologies and frameworks for learning behaviours and designs of robots with the goal of making them robust to mechanical failures or stark environmental changes impacting the performance of the robot. The student will initially perform a literature review, then formulate and implement an algorithm and test it in simulation. For further information, please contact the supervisor.

Minimum knowledge:

  • Good Python Skills
  • Basic Linux skills
  • Some prior experience with Deep Learning

Preferred knowledge:

  • Reinforcement Learning & Deep Reinforcement Learning

Thesis student will:

  • Get an overview over deep reinforcement learning algorithms and their application to practical problems
  • Learn to utilize and combine optimization techniques with machine learning algorithms
  • Work on state-of-the-art problems in research

Related Literature: