Saved in:
Bibliographic Details
Main Authors: Nai, Ruiqian, You, Jiacheng, Cao, Liu, Cui, Hanchen, Zhang, Shiyuan, Xu, Huazhe, Gao, Yang
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2502.10956
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916617393274880
author Nai, Ruiqian
You, Jiacheng
Cao, Liu
Cui, Hanchen
Zhang, Shiyuan
Xu, Huazhe
Gao, Yang
author_facet Nai, Ruiqian
You, Jiacheng
Cao, Liu
Cui, Hanchen
Zhang, Shiyuan
Xu, Huazhe
Gao, Yang
contents Legged locomotion is not just about mobility; it also encompasses crucial objectives such as energy efficiency, safety, and user experience, which are vital for real-world applications. However, key factors such as battery power consumption and stepping noise are often inaccurately modeled or missing in common simulators, leaving these aspects poorly optimized or unaddressed by current sim-to-real methods. Hand-designed proxies, such as mechanical power and foot contact forces, have been used to address these challenges but are often problem-specific and inaccurate. In this paper, we propose a data-driven framework for fine-tuning locomotion policies, targeting these hard-to-simulate objectives. Our framework leverages real-world data to model these objectives and incorporates the learned model into simulation for policy improvement. We demonstrate the effectiveness of our framework on power saving for quadruped locomotion, achieving a significant 24-28\% net reduction in total power consumption from the battery pack at various speeds. In essence, our approach offers a versatile solution for optimizing hard-to-simulate objectives in quadruped locomotion, providing an easy-to-adapt paradigm for continual improving with real-world knowledge. Project page https://hard-to-sim.github.io/.
format Preprint
id arxiv_https___arxiv_org_abs_2502_10956
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Fine-Tuning Hard-to-Simulate Objectives for Quadruped Locomotion: A Case Study on Total Power Saving
Nai, Ruiqian
You, Jiacheng
Cao, Liu
Cui, Hanchen
Zhang, Shiyuan
Xu, Huazhe
Gao, Yang
Robotics
Legged locomotion is not just about mobility; it also encompasses crucial objectives such as energy efficiency, safety, and user experience, which are vital for real-world applications. However, key factors such as battery power consumption and stepping noise are often inaccurately modeled or missing in common simulators, leaving these aspects poorly optimized or unaddressed by current sim-to-real methods. Hand-designed proxies, such as mechanical power and foot contact forces, have been used to address these challenges but are often problem-specific and inaccurate. In this paper, we propose a data-driven framework for fine-tuning locomotion policies, targeting these hard-to-simulate objectives. Our framework leverages real-world data to model these objectives and incorporates the learned model into simulation for policy improvement. We demonstrate the effectiveness of our framework on power saving for quadruped locomotion, achieving a significant 24-28\% net reduction in total power consumption from the battery pack at various speeds. In essence, our approach offers a versatile solution for optimizing hard-to-simulate objectives in quadruped locomotion, providing an easy-to-adapt paradigm for continual improving with real-world knowledge. Project page https://hard-to-sim.github.io/.
title Fine-Tuning Hard-to-Simulate Objectives for Quadruped Locomotion: A Case Study on Total Power Saving
topic Robotics
url https://arxiv.org/abs/2502.10956