Attention-based map encoding for learning generalized legged locomotion
Science Robotics · 27.08.2025 · 10 (105)
DOI: 10.1126/scirobotics.adv3604 ↗Kurz gefasst
Fachartikel in der Zeitschrift Science Robotics, erschienen am 27.08.2025. Verfasst von Junzhe He und 5 weiteren Autor:innen. Beteiligte Einrichtung: Robotic Systems Lab, ETH Zurich, 8092 Zurich, Switzerland. Die Arbeit wurde bisher 31-mal zitiert (Crossref) und nennt 49 Literaturangaben.
Abstract (Originalsprache)
Dynamic locomotion of legged robots is a critical yet challenging topic in expanding the operational range of mobile robots. It requires precise planning when possible footholds are sparse, robustness against uncertainties and disturbances, and generalizability across diverse terrains. Although traditional model-based controllers excel at planning on complex terrains, they struggle with real-world uncertainties. Learning-based controllers offer robustness to such uncertainties but often lack precision on terrains with sparse steppable areas. Hybrid methods achieve enhanced robustness on sparse terrains by combining both methods but are computationally demanding and constrained by the inherent limitations of model-based planners. To achieve generalized legged locomotion on diverse terrains while preserving the robustness of learning-based controllers, this paper proposes an attention-based map encoding conditioned on robot proprioception, which is trained as part of the controller using reinforcement learning. We show that the network learns to focus on steppable areas for future footholds when the robot dynamically navigates diverse and challenging terrains. We synthesized behaviors that exhibited robustness against uncertainties while enabling precise and agile traversal of sparse terrains. In addition, our method offers a way to interpret the topographical perception of a neural network. We have trained two controllers for a 12-degrees-of-freedom quadrupedal robot and a 23-degrees-of-freedom humanoid robot and tested the resulting controllers in the real world under various challenging indoor and outdoor scenarios, including ones unseen during training.
Autor:innen und Institutionen
- Robotic Systems Lab, ETH Zurich, 8092 Zurich, Switzerland.
- Disney Research Zurich, Stampfenbachstrasse 48, 8006 Zurich, Switzerland.
ORCID-Profile: Junzhe He, Chong Zhang, Fabian Jenelten, Ruben Grandia, Moritz Bächer, Marco Hutter
Zitieren
APA
He, J., Zhang, C., Jenelten, F., Grandia, R., Bächer, M., & Hutter, M. (2025). Attention-based map encoding for learning generalized legged locomotion. Science Robotics, 10(105). https://doi.org/10.1126/scirobotics.adv3604
Vancouver
He J, Zhang C, Jenelten F, Grandia R, Bächer M, Hutter M. Attention-based map encoding for learning generalized legged locomotion. Science Robotics. 2025;10(105). doi:10.1126/scirobotics.adv3604
BibTeX
@article{He2025,
author = {He, Junzhe and Zhang, Chong and Jenelten, Fabian and Grandia, Ruben and Bächer, Moritz and Hutter, Marco},
title = {Attention-based map encoding for learning generalized legged locomotion},
journal = {Science Robotics},
year = {2025},
volume = {10},
number = {105},
doi = {10.1126/scirobotics.adv3604}
}
EUVIMED