RAIBO2, KAIST’s quadruped robot dog , has completed an official 42.195-kilometer marathon at the Sangju Dried Persimmon Marathon in South Korea. The research team, led by Professor Je Min Hwangbo, achieved this feat by integrating reinforcement learning algorithms into their proprietary simulator, RaiSim, which prepared the robot for challenging terrain such as slopes and icy surfaces.
Quadruped robot faces and overcomes challenges
The marathon course presented significant challenges, including steep 50-meter climbs. However, RAIBO2’s high-torque articulation mechanism allowed it to efficiently harvest energy when descending hills, maximizing the use of energy resources.
Furthermore, the collaboration with RAION Robotics Inc. was essential, increasing the robot ‘s stability and ensuring its ability to operate for more than 4 hours in an environment full of people and obstacles.
PhD candidate Choongin Lee highlighted that this project demonstrates RAIBO2 ’s potential to perform functions in urban environments, such as patrolling and delivery , and indicated that the team is working on autonomous navigation functions to tackle even greater challenges, such as disasters and mountainous terrain. This development was made possible with support from Samsung Electronics’ Future Technology Promotion Center, underscoring how innovation continues to redefine the boundaries of robotics .
KAIST quadruped robot in the marathon. Source: MBN
Follow us on social media and don’t miss any of our posts!
YouTube LinkedIn Facebook Instagram X TikTok
Source and photos: KAIST