Learning Locomotion on Complex Terrain for Quadrupedal Robots with Foot Position Maps and Stability Rewards

Matthew Hwang, Yubin Liu, Ryo Hakoda Takeshi Oishi
The University of Tokyo
Paper Code arXiv

Abstract

Quadrupedal locomotion over complex terrain has been a long-standing research topic in robotics. While recent reinforcement learning-based locomotion methods improve generalizability and foot-placement precision, they rely on implicit inference of foot positions from joint angles, lacking the explicit precision and stability guarantees of optimization-based approaches. To address this, we introduce a foot position map integrated into the heightmap, and a dynamic locomotion-stability reward within an attention-based framework to achieve locomotion on complex terrain. We validate our method extensively on terrains seen during training as well as out-of-domain (OOD) terrains. Our results demonstrate that the proposed method enables precise and stable movement, resulting in improved locomotion success rates on both in-domain and OOD terrains.

Method Pipeline

Method Pipeline

Locomotion Video

BibTeX

@misc{hwang2026learninglocomotioncomplexterrain,
      title={Learning Locomotion on Complex Terrain for Quadrupedal Robots with Foot Position Maps and Stability Rewards}, 
      author={Matthew Hwang and Yubin Liu and Ryo Hakoda and Takeshi Oishi},
      year={2026},
      eprint={2604.02744},
      archivePrefix={arXiv},
      primaryClass={cs.RO},
      url={https://arxiv.org/abs/2604.02744}, 
}