TAPNAV: Humanoid Navigation through Tactile Active Perception

2026-10-07

AI Summary

This study addresses the navigation challenges faced by humanoid robots in vision-denied environments due to odometry drift, proposing a visionless navigation framework based on active tactile perception. The method pioneers the use of deliberate physical interactions to acquire localization cues, achieving reliable navigation by coupling uncertainty-aware global planning with information-gain-driven local exploration. Furthermore, it fuses multi-source observations from IMU, odometry, and tactile sensing, while employing a whole-body controller to coordinate locomotion and contact. Simulation and real-world experiments on the Unitree G1 platform demonstrate that the proposed framework significantly reduces state estimation errors and improves task completion rates. This work establishes a novel paradigm for autonomous robot navigation in extreme environments where visual feedback is unavailable.

Abstract

Navigation in vision-denied environments is challenging for humanoid robots because proprioceptive odometry drifts and localization uncertainty accumulates rapidly. We present TAPNAV, a tactile active-perception framework that enables humanoid navigation toward a goal by actively probing surrounding structures without relying on vision. TAPNAV maintains a pose belief from odometry, IMU, and tactile contact observations, and couples uncertainty-aware global route planning with information-gain-driven local probing. The global planner searches for routes that keep predicted localization uncertainty bounded by exploiting opportunities for tactile correction, while the local planner selects probe actions that maximize expected information gain. A whole-body controller coordinates the humanoid's locomotion and end-effector contact to execute the planned navigation and probe motions. We evaluate TAPNAV in simulation and on a Unitree G1 across different floor plans and obstacle geometries. TAPNAV achieves lower state estimation error and a higher task completion rate than baselines. These results demonstrate that actively planning physical interactions with the environment can provide localization cues for reliable humanoid navigation without vision.

Problem

• Humanoid Navigation
• Vision-denied Environments
• Tactile Perception
• Localization Uncertainty
• Odometry Drift

Innovation

• Tactile Active Perception
• Humanoid Navigation
• Uncertainty-aware Planning
• Information Gain
• Whole-body Control