A New AI Controller Helps Humanoid Robots Think on Their Feet
Overview
Researchers at ETH Zurich have developed a unified AI control system for humanoid robots that enables versatile locomotion tasks such as climbing ladders, vaulting over obstacles, and recovering from falls. Using an attention-based contact planning approach combined with motion imitation and reinforcement learning, the system identifies optimal contact points for nine body parts and coordinates 23 degrees of freedom to navigate complex terrains. Demonstrated on a Unitree G1 robot, the controller achieved a 92.5% success rate in simulation across eight challenging obstacle courses, compared to 58.8% for a competing method. The system operates at 50 Hz using onboard LiDAR for real-time terrain adaptation, though it currently focuses on local navigation and terrain-based interactions rather than global autonomy.
Generated from attributed reports · Updated 13 minutes ago
Event evidence and corrections
0 attributed source owners. Ownership does not establish independent confirmation. Quantities are reported separately and are never added together.
Reported quantity · Reported success rate: 92.5 percent · reported trials
“uously adapting movements to the surrounding environment. In simulation, the new system achieved a 92.5% success rate across eight challenging obstacle courses, compared with 58.8% for a competing motion-generation approach. Real-world demonstrations also showed the robot recovering from failed jumps, unsuccessful attempts to stand, and getting stuck while climbing a ladder. Importantly, these behaviors emerge from a”
Exact source · revision 2Source owner not reported
Reported quantity · number of obstacle courses in simulation: 8 other · reported trials
“uously adapting movements to the surrounding environment. In simulation, the new system achieved a 92.5% success rate across eight challenging obstacle courses, compared with 58.8% for a competing motion-generation approach. Real-world demonstrations also showed the robot recovering from failed jumps, unsuccessful attempts to stand, and getting stuck while climbing a ladder. Importantly, these behaviors emerge from a”
Exact source · revision 2Source owner not reported
Reported quantity · frequency of controller operation: 50 other · Basis not reported
“ct useful contact locations based on terrain observations, the robot's physical state, and its destination. Finally, reinforcement learning allows the system to refine these abilities on unfamiliar terrain. At runtime, onboard LiDAR sensors provide detailed terrain information, allowing the robot to identify features as small as ladder rungs. Both the contact planner and whole-body controller operate at 50 Hz, contin”
Exact source · revision 2Source owner not reported
Reported quantity · number of body parts for contact planning: 9 other · Basis not reported
“fies suitable contact points for nine body parts, including the hands, elbows, knees, feet, and pelvis. A separate controller then coordinates the robot's 23 actuated degrees of freedom to execute those plans. Training the system takes place in three stages. First, the robot learns coordinated movements from human motion data adapted to its mechanical structure. Next, an attention-based neural network learns to predi”
Exact source · revision 2Source owner not reported
Reported quantity · number of actuated degrees of freedom: 23 other · Basis not reported
“fies suitable contact points for nine body parts, including the hands, elbows, knees, feet, and pelvis. A separate controller then coordinates the robot's 23 actuated degrees of freedom to execute those plans. Training the system takes place in three stages. First, the robot learns coordinated movements from human motion data adapted to its mechanical structure. Next, an attention-based neural network learns to predi”
Exact source · revision 2Source owner not reported
Reported quantity · training stages: 3 other · Basis not reported
“fies suitable contact points for nine body parts, including the hands, elbows, knees, feet, and pelvis. A separate controller then coordinates the robot's 23 actuated degrees of freedom to execute those plans. Training the system takes place in three stages. First, the robot learns coordinated movements from human motion data adapted to its mechanical structure. Next, an attention-based neural network learns to predi”
Exact source · revision 2Source owner not reported
Report timeline
Follow attributed reports and material updates.
- Robotics — Robot learning news discoverySignalA New AI Controller Helps Humanoid Robots Think on Their Feet
Researchers at ETH Zurich have developed a unified AI control system for humanoid robots that enables versatile locomotion tasks such as climbing ladders, vaulting over obstacles, and recovering from falls. Using an attention-based contact planning approach combined with motion imitation and reinforcement learning, the system identifies optimal contact points for nine body parts and coordinates 23 degrees of freedom to navigate complex terrains. Demonstrated on a Unitree G1 robot, the controller achieved a 92.5% success rate in simulation across eight challenging obstacle courses, compared to 58.8% for a competing method. The system operates at 50 Hz using onboard LiDAR for real-time terrain adaptation, though it currently focuses on local navigation and terrain-based interactions rather than global autonomy.
Event coverage history
There is not enough continuous observation data to show a trend.
Timezone · UTC
Article dates follow your selected timezone. Briefing editions use Hong Kong time (UTC+8).