Reinforcement Learning for Robust Athletic Intelligence: Lessons from the 2nd 'AI Olympics with RealAIGym' Competition
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910884130979840 |
|---|---|
| author | Wiebe, Felix Turcato, Niccolò Libera, Alberto Dalla Choe, Jean Seong Bjorn Choi, Bumkyu Faust, Tim Lukas Maraqten, Habib Aghadavoodi, Erfan Cali, Marco Sinigaglia, Alberto Giacomuzzo, Giulio Romeres, Diego Kim, Jong-kook Susto, Gian Antonio Vyas, Shubham Mronga, Dennis Belousov, Boris Peters, Jan Kirchner, Frank Kumar, Shivesh |
| author_facet | Wiebe, Felix Turcato, Niccolò Libera, Alberto Dalla Choe, Jean Seong Bjorn Choi, Bumkyu Faust, Tim Lukas Maraqten, Habib Aghadavoodi, Erfan Cali, Marco Sinigaglia, Alberto Giacomuzzo, Giulio Romeres, Diego Kim, Jong-kook Susto, Gian Antonio Vyas, Shubham Mronga, Dennis Belousov, Boris Peters, Jan Kirchner, Frank Kumar, Shivesh |
| contents | In the field of robotics many different approaches ranging from classical planning over optimal control to reinforcement learning (RL) are developed and borrowed from other fields to achieve reliable control in diverse tasks. In order to get a clear understanding of their individual strengths and weaknesses and their applicability in real world robotic scenarios is it important to benchmark and compare their performances not only in a simulation but also on real hardware. The '2nd AI Olympics with RealAIGym' competition was held at the IROS 2024 conference to contribute to this cause and evaluate different controllers according to their ability to solve a dynamic control problem on an underactuated double pendulum system with chaotic dynamics. This paper describes the four different RL methods submitted by the participating teams, presents their performance in the swing-up task on a real double pendulum, measured against various criteria, and discusses their transferability from simulation to real hardware and their robustness to external disturbances. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_15290 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Reinforcement Learning for Robust Athletic Intelligence: Lessons from the 2nd 'AI Olympics with RealAIGym' Competition Wiebe, Felix Turcato, Niccolò Libera, Alberto Dalla Choe, Jean Seong Bjorn Choi, Bumkyu Faust, Tim Lukas Maraqten, Habib Aghadavoodi, Erfan Cali, Marco Sinigaglia, Alberto Giacomuzzo, Giulio Romeres, Diego Kim, Jong-kook Susto, Gian Antonio Vyas, Shubham Mronga, Dennis Belousov, Boris Peters, Jan Kirchner, Frank Kumar, Shivesh Robotics In the field of robotics many different approaches ranging from classical planning over optimal control to reinforcement learning (RL) are developed and borrowed from other fields to achieve reliable control in diverse tasks. In order to get a clear understanding of their individual strengths and weaknesses and their applicability in real world robotic scenarios is it important to benchmark and compare their performances not only in a simulation but also on real hardware. The '2nd AI Olympics with RealAIGym' competition was held at the IROS 2024 conference to contribute to this cause and evaluate different controllers according to their ability to solve a dynamic control problem on an underactuated double pendulum system with chaotic dynamics. This paper describes the four different RL methods submitted by the participating teams, presents their performance in the swing-up task on a real double pendulum, measured against various criteria, and discusses their transferability from simulation to real hardware and their robustness to external disturbances. |
| title | Reinforcement Learning for Robust Athletic Intelligence: Lessons from the 2nd 'AI Olympics with RealAIGym' Competition |
| topic | Robotics |
| url | https://arxiv.org/abs/2503.15290 |