Reinforcement learning enables quadruped robot manipulation on rough terrain
Reinforcement Learning Based Multi-Terrain Adaptive Whole-Body Control for a Quadruped Manipulator
Robotics
Summary
Walking robots with arms can do jobs in tricky places, but it's hard to keep their arms steady while moving over uneven ground. The authors designed a computer learning method that lets a small four-legged robot control its whole body to handle tough surfaces like hills or stairs. They built a lightweight robot and tested this method in real-world settings, showing it works well. This approach helps the robot’s arm stay on task while it walks over bumpy or shifting terrain.
What this means in practice
- •For robotics engineers: Enable four-legged robots to perform stable manipulation tasks while walking on uneven and deformable surfaces using learned whole-body control.
- •For disaster response teams: Deploy small quadruped robots capable of manipulating objects in complex terrain such as slopes and stairs for search and rescue tasks.
Authors
Yanzhou Jin, Fei Suo, Chang Liu, Yasuhisa Hasegawa, Liming Chen, Yusuke Iwasawa, Yutaka Matsuo, Yaonan Zhu
Abstract
Legged robots equipped with manipulators offer promising capabilities for loco-manipulation tasks in unstructured environments. However, achieving task-consistent end-effector behavior during locomotion remains challenging due to terrain-induced disturbances and the strong coupling between locomotion and manipulation, particularly for small-scale quadruped platforms. A reinforcement-learning based whole-body control framework is proposed for quadruped loco-manipulation on complex terrain. The proposed framework combines a hybrid task representation with task-prioritized whole-body learning to coordinate locomotion and manipulation while preserving task-consistent end-effector behavior. A lightweight quadruped manipulator platform is further developed to validate the proposed framework on a small-scale quadruped system. Real-world experiments demonstrate successful deployment on complex terrains such as slopes up to 30 degrees, stairs, and deformable terrain. These results demonstrate the feasibility of deploying learned whole-body control policy for quadruped loco-manipulation in real-world environments and highlight the importance of task-consistent whole-body coordination for terrain-adaptive loco-manipulation.