GuideWalk: Learning Unified Autonomous Navigation and Locomotion for Humanoid Robots across Versatile Terrains

2026-06-09

AI Summary

This work addresses the challenge of achieving reliable navigation and dynamically feasible locomotion for humanoid robots in complex terrains. To this end, the authors propose GuideWalk, an end-to-end unified framework that decouples obstacle avoidance from terrain adaptation by integrating traversability-aware navigation guidance with terrain-adaptive motion policies. The core innovations include an explicit velocity guidance module and a composite teacher distillation mechanism, which effectively combine goal-oriented commands with dynamically consistent actions. The policy is jointly optimized through reinforcement learning and behavioral cloning. Experimental results demonstrate that GuideWalk enables stable, efficient, and robust navigation and walking for humanoid robots across diverse environments.

Abstract

Humanoid robots have achieved strong locomotion capabilities, but reliable navigation on versatile terrains remains challenging because obstacle avoidance must be coordinated with dynamically feasible motion. In this work, we present GuideWalk, a unified end-to-end framework that integrates traversability-aware navigation guidance with terrain-adaptive locomotion teacher for humanoid navigation. Specifically, we introduce a navigation module that provides explicit velocity guidance, decoupling obstacle avoidance from terrain conditions to enable robust planning across diverse environments. We propose a composite teacher distillation scheme, where goal-directed commands and dynamically consistent actions are aggregated and distilled into a single policy. To further improve robustness, the distilled policy is refined with reinforcement learning and an auxiliary behavior cloning objective, which promotes exploration while preserving desirable teacher behaviors. Experiments demonstrate that GuideWalk achieves stable and effective navigation while maintaining stable humanoid locomotion.

Problem

• humanoid robots
• autonomous navigation
• versatile terrains
• locomotion
• obstacle avoidance

Innovation

• unified navigation and locomotion
• traversability-aware guidance
• composite teacher distillation
• humanoid robots
• end-to-end learning