Reinforcement Learning Group and Shanghai Banxing Robotics Successfully Complete the 2025 Beijing Humanoid Robot Half Marathon
On April 19, 2025, the Reinforcement Learning Group (RLG), in collaboration with Shanghai Banxing Robotics Co., Ltd., successfully participated in the 2025 Beijing Yizhuang Humanoid Robot Half Marathon. As one of the first large-scale competitions focused on humanoid robot endurance and mobility, the event attracted top institutions and companies from across the country.
In this joint effort, RLG was responsible for developing and implementing the robot’s control and locomotion algorithms, while Banxing Robotics provided the hardware platform and technical support. The humanoid robot, co-developed by the two teams, completed the full half marathon distance and achieved fifth place overall, finishing with a total time of 4 hours, 2 minutes, and 19 seconds.
This successful completion marks a significant milestone in advancing long-distance locomotion for humanoid robots. It demonstrates the effectiveness of combining model-based predictive control and reinforcement learning techniques with a robust and reliable hardware system under realistic, dynamic conditions.
