Quadruped robots estimate forces for better walking and leash control
Proprioceptive Force Estimation for Quadruped Locomotion and Human-Robot Interaction
Robotics
Summary
Walking robots that carry loads or are guided by leashes need to understand forces acting on them. The authors designed a way for four-legged robots to estimate these forces using only their own sensors and past movements. This helps the robot walk more stably with heavy loads and respond to leash pulls to move as desired. They tested their approach both in simulation and on a real robot, showing improved balance and control compared to previous methods.
What this means in practice
- •For robotics engineers: Improve quadruped robot stability and control by integrating force estimation from internal sensors during payload carrying and leash-guided motion.
- •For robot product developers: Develop quadruped delivery or service robots that keep balance under varying loads and follow human commands via leash force sensing.$Commercial implications: Enables sale of quadruped robots capable of robust load carrying and intuitive human interaction through leash guidance.
Authors
Run Wang, Xu Yang, Alapati Tuerxun, Yilin Mo
Abstract
Payload forces must be accommodated during locomotion, while leash forces can specify desired motion. We investigate whether a shared three-dimensional force estimate in newtons, inferred from proprioceptive history under sustained loading, can support both tasks. An estimator and locomotion policy are jointly trained with supervised force and velocity outputs and learned latent context. The estimated force conditions locomotion and additionally generates planar-velocity and yaw-rate commands for leash guidance through an analytical map. In sustained-force simulation sweeps, temporal means of componentwise force root mean square error range from 1.44 to 2.83\,N. Compared with a domain-randomized baseline, the framework reduces velocity-tracking and base-orientation error scores by 21.6\% and 46.5\%, respectively, and increases mean survival from 68.29\% to 94.60\% in separate sustained-force tests. Unitree Go1 experiments demonstrate stationary vertical and horizontal force estimation, locomotion with an 8.5\,kg payload whose weight exceeds the 70\,N training force limit, and leash guidance using the same force-estimation interface.