Abstract
Unknown dynamic load carrying is one important practical application for quadruped robots. Such a problem is non-trivial, posing three major challenges in quadruped loco-motion control. First, how to model or represent the dynamics of the load in a generic manner. Second, how to make the robot capture the dynamics without any external sensing. Third, how to enable the robot to interact with load handling the mutual effect and stabilizing the load. In this work, we propose a general load modeling approach called load characteristics modeling to capture the dynamics of the load. We integrate this proposed modeling technique and leverage recent advances in Reinforcement Learning (RL) based locomotion control to enable the robot to infer the dynamics of load movement and interact with the load indirectly to stabilize it. We conduct extensive comparative simulation experiments to validate the effectiveness and superiority of our proposed method. Results show that our method outperforms other methods in sudden load resistance, load stabilizing and locomotion with heavy load on rough terrain.
Video
Method
We propose a general load modeling approach called load characteristics modeling to capture the dynamics of the load, consisting of [load mass, load friction coefficient, load position, load velocity]. We integrate this proposed modeling technique and leverage recent advances in Reinforcement Learning (RL) based locomotion control to enable the robot to infer the dynamics of load movement and interact with the load indirectly to stabilize it.
We train a teacher policy netowrk with the load characteristics as privileged infomation and concurrently train a load estimator to predict the load characteristics from historical proprioception. And the teacher-student distillation technique here is employed to reconstruct the compressed latent state from proprioception history to enable the sim-to-real deployment of blind locomotion.
Simulation Experiments
Comparison of different methods on rough terrain with 6kg load and 0.01 friction coefficient.
Baseline
NLW
LW
Ours
BibTeX
@article{chang2025leggedloadadapt,
author = {Chang, Leixin and Nai, Yuxuan and Chen, Hua and Yang, Liangjing},
title = {Beyond Robustness: Learning Unknown Dynamic Load Adaptation for Quadruped Locomotion on Rough Terrain},
journal = {ICRA},
year = {2025},
}