Collision-Free Humanoid Traversal in Cluttered Indoor Scenes

2026-01-22

AI Summary

This work addresses the challenge of enabling humanoid robots to navigate cluttered indoor environments without collisions—such as stepping over obstacles, ducking under low structures, or squeezing through narrow passages—by introducing HumanoidPF, a novel potential field representation that maps environmental perception to feasible motion directions, thereby substantially reducing the complexity of training reinforcement learning policies. Leveraging a hybrid clutter generation strategy combining real-world 3D scene crops with procedurally generated obstacles, and an efficient sim-to-real transfer pipeline, the approach achieves high success rates in both simulation and physical deployment. Furthermore, the system integrates an innovative single-click teleoperation interface, allowing users to guide the robot through complex indoor scenes with just one click, demonstrating strong practicality and generalization capability.

Technology Category

• Intelligent Robots: Human-Robot Interaction
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Machine Learning: Imitation Learning & Inverse Reinforcement Learning

Application Category

• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications

Abstract

We study the problem of collision-free humanoid traversal in cluttered indoor scenes, such as hurdling over objects scattered on the floor, crouching under low-hanging obstacles, or squeezing through narrow passages. To achieve this goal, the humanoid needs to map its perception of surrounding obstacles with diverse spatial layouts and geometries to the corresponding traversal skills. However, the lack of an effective representation that captures humanoid-obstacle relationships during collision avoidance makes directly learning such mappings difficult. We therefore propose Humanoid Potential Field (HumanoidPF), which encodes these relationships as collision-free motion directions, significantly facilitating RL-based traversal skill learning. We also find that HumanoidPF exhibits a surprisingly negligible sim-to-real gap as a perceptual representation. To further enable generalizable traversal skills through diverse and challenging cluttered indoor scenes, we further propose a hybrid scene generation method, incorporating crops of realistic 3D indoor scenes and procedurally synthesized obstacles. We successfully transfer our policy to the real world and develop a teleoperation system where users could command the humanoid to traverse in cluttered indoor scenes with just a single click. Extensive experiments are conducted in both simulation and the real world to validate the effectiveness of our method. Demos and code can be found in our website: https://axian12138.github.io/CAT/.

Problem

• collision-free
• humanoid traversal
• cluttered indoor scenes
• obstacle avoidance
• motion planning

Innovation

• Humanoid Potential Field
• collision-free traversal
• sim-to-real transfer
• hybrid scene generation
• reinforcement learning