TR2024-108

Reinforcement Learning for Athletic Intelligence: Lessons from the 1st “AI Olympics with RealAIGym” Competition

    •  Wiebe, F., Turcato, N., Dalla Libera, A., Zhang, C., Vincent, T., Vyas, S., Giacomuzzo, G., Carli, R., Romeres, D., Sathuluri, A., Zimmermann, M., Belousov, B., Peters, J., Kirchner, F., Kumar, S., "Reinforcement Learning for Athletic Intelligence: Lessons from the 1st “AI Olympics with RealAIGym” Competition", Internation Joint Conference on Artificial Intelligence (IJCAI), August 2024.
      BibTeX TR2024-108 PDF
      • @inproceedings{Wiebe2024aug,
      • author = {Wiebe, Felix and Turcato, Niccolò and Dalla Libera, Alberto and Zhang, Chi and Vincent, Theo and Vyas, Shubham and Giacomuzzo, Giulio and Carli, Ruggero and Romeres, Diego and Sathuluri, Akhil and Zimmermann, Markus and Belousov, Boris and Peters, Jan and Kirchner, Frank and Kumar, Shivesh}},
      • title = {Reinforcement Learning for Athletic Intelligence: Lessons from the 1st “AI Olympics with RealAIGym” Competition},
      • booktitle = {Internation Joint Conference on Artificial Intelligence (IJCAI)},
      • year = 2024,
      • month = aug,
      • url = {https://www.merl.com/publications/TR2024-108}
      • }
  • MERL Contact:
  • Research Area:

    Robotics

Abstract:

As artificial intelligence gains new capabilities, it becomes important to evaluate it on real-world tasks. In particular, the fields of robotics and reinforcement learning (RL) are lacking in standardized benchmarking tasks on real hardware. To facilitate reproducibility and stimulate algorithmic advancements, we held an AI Olympics competi- tion at IJCAI 2023 conference based on the double pendulum system in the RealAIGym project where the participants were asked to develop a controller for the swing up and stabilization task. This paper presents the methods and results from the top par- ticipating teams and provides insights into the real- world performance of RL algorithms with respect to a baseline time-varying LQR controller.