Learning and Verifying Complex Behaviour of Humanoid Robots

Learning and Verifying Complex Behaviour of Humanoid Robots

  • Duration:
  • Application fields

The validation of systems based on deep learning for use in safety-critical applications proves to be inherently difficult, since their subsymbolic mode of operation does not provide adequate levels of abstraction for representation and proof of correctness. The VeryHuman project aims to synthesize such levels of abstraction by observing and analysing the behaviour of upright walking of a two-legged humanoid robot. The theory to be developed is the starting point for the definition of an appropriate reward function to optimally control the movements of the humanoid by means of enhanced learning, as well as for verifiable abstraction of the corresponding kinematic models, which can be used to validate the behaviour of the robot more easily.


Cyber Physical Systems (CPS), DFKI Robotics Innovation Center (RIC), DFKI


Federal Ministry of Education and Research (BMBF)


Federal Ministry of Education and Research (BMBF)


[Translate to English:]

Publications about the project

Moritz Schilling, Shivesh Kumar, Heiner Peters, Adriano del Rio Fernandez, Christoph Stoeffler

In: Bruno Siciliano, Oussama Khatib (editor). ARK 2022: Advances in Robot Kinematics 2022. International Symposium on Advances in Robot Kinematics (ARK-2022) June 27-30 Bilbao Spain Pages 348-355 Springer Proceedings in Advanced Robotics (SPAR) 24 ISBN 978-3-031-08140-8 Springer Cham 6/2022.

To the publication

German Research Center for Artificial Intelligence
Deutsches Forschungszentrum für Künstliche Intelligenz