Humanoid robot control improves collision safety with scene awareness
Collision-Aware Humanoid Whole-Body Control under Imperfect Tracking Targets
Robotics
Summary
Humanoid robots move their whole bodies to perform tasks, but their controllers often don’t see the environment around them, which can cause crashes. The authors created a method called RECAL that helps the robot avoid bumping into things by looking at a 3D map of the robot, objects it’s holding, and the surroundings. This method balances following the robot’s movement targets while keeping it from hitting obstacles. Their tests in simulations and on a real robot show better safety without losing accuracy in movement.
What this means in practice
- •For robotic system engineers: Develop controllers for humanoid robots that prevent collisions by integrating 3D environment data with motion commands.
- •For warehouse automation teams: Enhance humanoid robot deployments in cluttered spaces by enabling safer navigation and object handling that reduces crash risks.$Commercial implications: Creates more reliable robotic handlers for commercial warehouses by minimizing damage from collisions in busy environments.
Authors
Mohitvishnu S. Gadde, Ashish Malik, Pranay Dugar, Aayam Kumar Shrestha, Alan Fern
Abstract
Humanoid robots often execute motion commands through whole-body controllers (WBCs) that track targets while maintaining balance and stability. However, most WBCs are blind to scene geometry, which can lead to collisions from imperfect target motions that are geometrically unsafe due to perception, planning, or teleoperation errors. We propose RECAL, a Robot--Environment Cross-Attention Layer that wraps a blind WBC to trade off target tracking against collision avoidance using external scene geometry. RECAL supports collision-aware tracking of floating-base and end-effector commands, including collision avoidance for held objects. It represents the robot, held objects, and environment as point clouds, using cross-attention between robot/object points and the environment to produce geometry-aware control features. In simulation, RECAL improves collision avoidance while preserving target-tracking performance across frozen-arm and adaptive-arm locomotion, object-carrying, and standing-manipulation scenarios relative to alternative geometry-aware WBC architectures. We further demonstrate the controller on a real Digit V3 humanoid robot.