Hiking in the Wild: A Scalable Perceptive Parkour Framework for Humanoids

TL;DR

This paper introduces 'Hiking in the Wild', a scalable end-to-end reinforcement learning framework enabling humanoid robots to traverse complex terrains safely and efficiently using raw depth data and proprioception.

Contribution

It presents a novel perceptive parkour framework with safety mechanisms and sampling strategies, eliminating reliance on external state estimation for robust humanoid hiking.

Findings

Enables humanoid robots to hike at speeds up to 2.5 m/s in complex terrains.

Introduces safety and sampling mechanisms to improve training stability and safety.

Open-sourced code facilitates reproducibility and deployment on real robots.

Abstract

Achieving robust humanoid hiking in complex, unstructured environments requires transitioning from reactive proprioception to proactive perception. However, integrating exteroception remains a significant challenge: mapping-based methods suffer from state estimation drift; for instance, LiDAR-based methods do not handle torso jitter well. Existing end-to-end approaches often struggle with scalability and training complexity; specifically, some previous works using virtual obstacles are implemented case-by-case. In this work, we present \textit{Hiking in the Wild}, a scalable, end-to-end parkour perceptive framework designed for robust humanoid hiking. To ensure safety and training stability, we introduce two key mechanisms: a foothold safety mechanism combining scalable \textit{Terrain Edge Detection} with \textit{Foot Volume Points} to prevent catastrophic slippage on edges, and a \textit{Flat Patch Sampling} strategy that mitigates reward hacking by generating feasible navigation targets. Our approach utilizes a single-stage reinforcement learning scheme, mapping raw depth inputs and proprioception directly to joint actions, without relying on external state estimation. Extensive field experiments on a full-size humanoid demonstrate that our policy enables robust traversal of complex terrains at speeds up to 2.5 m/s. The training and deployment code is open-sourced to facilitate reproducible research and deployment on real robots with minimal hardware modifications.