Saved in:
Bibliographic Details
Main Authors: Mishra, Prakhar, Raj, Amir Hossain, Xiao, Xuesu, Manocha, Dinesh
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2505.18429
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912715786682368
author Mishra, Prakhar
Raj, Amir Hossain
Xiao, Xuesu
Manocha, Dinesh
author_facet Mishra, Prakhar
Raj, Amir Hossain
Xiao, Xuesu
Manocha, Dinesh
contents We address the problem of agile and rapid locomotion, a key characteristic of quadrupedal and bipedal robots. We present a new algorithm that maintains stability and generates high-speed trajectories by considering the temporal aspect of locomotion. Our formulation takes into account past information based on a novel history-aware curriculum Learning (HACL) algorithm. We model the history of joint velocity commands with respect to the observed linear and angular rewards using a recurrent neural net (RNN). The hidden state helps the curriculum learn the relationship between the forward linear velocity and angular velocity commands and the rewards over a given time-step. We validate our approach on the MIT Mini Cheetah,Unitree Go1, and Go2 robots in a simulated environment and on a Unitree Go1 robot in real-world scenarios. In practice, HACL achieves peak forward velocity of 6.7 m/s for a given command velocity of 7m/s and outperforms prior locomotion algorithms by nearly 20%.
format Preprint
id arxiv_https___arxiv_org_abs_2505_18429
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle HACL: History-Aware Curriculum Learning for Fast Locomotion
Mishra, Prakhar
Raj, Amir Hossain
Xiao, Xuesu
Manocha, Dinesh
Robotics
We address the problem of agile and rapid locomotion, a key characteristic of quadrupedal and bipedal robots. We present a new algorithm that maintains stability and generates high-speed trajectories by considering the temporal aspect of locomotion. Our formulation takes into account past information based on a novel history-aware curriculum Learning (HACL) algorithm. We model the history of joint velocity commands with respect to the observed linear and angular rewards using a recurrent neural net (RNN). The hidden state helps the curriculum learn the relationship between the forward linear velocity and angular velocity commands and the rewards over a given time-step. We validate our approach on the MIT Mini Cheetah,Unitree Go1, and Go2 robots in a simulated environment and on a Unitree Go1 robot in real-world scenarios. In practice, HACL achieves peak forward velocity of 6.7 m/s for a given command velocity of 7m/s and outperforms prior locomotion algorithms by nearly 20%.
title HACL: History-Aware Curriculum Learning for Fast Locomotion
topic Robotics
url https://arxiv.org/abs/2505.18429