QuietWalk: Physics-Informed Reinforcement Learning for Ground Reaction Force-Aware Humanoid Locomotion Under Diverse Footwear

2026-04-26

AI Summary

This work addresses the challenge of impact-induced vibrations and noise during humanoid robot locomotion in domestic and clinical settings, which degrade user experience, accelerate hardware wear, and hinder control generalization due to varying footwear-induced contact dynamics. To mitigate these issues, the authors propose QuietWalk, a novel framework that, for the first time, embeds inverse dynamics consistency constraints into a physics-informed neural network (PINN) to accurately estimate vertical ground reaction forces (GRF) using only proprioceptive signals. This GRF estimate is then incorporated as a penalty term in reinforcement learning to achieve low-impact gait control without force sensors. Experiments demonstrate an 82%–86% reduction in GRF prediction error (R² = 0.99) on real robot data, along with a 7.17 dB average and 4.98 dB peak reduction in A-weighted noise at 1.2 m/s, while maintaining robust performance across four footwear types and multiple floor surfaces.

Technology Category

• Intelligent Robots: Motion and Path Planning
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Natural Language Processing: Safety and Robustness

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Humanoid robots operating in human-centered environments (e.g., homes, hospitals, and offices) must mitigate foot--ground impact transients, as impact-induced vibration and noise degrade user experience and repeated impacts accelerate hardware wear. However, existing low-noise locomotion training often relies on kinematic proxy objectives or fragile force sensors, and footwear-induced changes in contact dynamics introduce distribution shifts that hinder policy generalization.We present QuietWalk, a physics-informed reinforcement learning framework for ground-reaction-force-aware humanoid locomotion under diverse footwear conditions. QuietWalk employs an inverse-dynamics-constrained physics-informed neural network (PINN) to estimate per-foot vertical ground reaction forces (GRFs) from proprioceptive signals, and integrates the frozen predictor into the RL training loop to penalize predicted impact forces without requiring force sensors at deployment.On a held-out real-robot dataset, enforcing inverse-dynamics consistency reduces vertical GRF prediction errors by 82%-86% compared with a purely supervised predictor and improves the coefficient of determination from 0.39/0.67 to 0.99/0.99 for the left/right feet. On hardware at 1.2 m/s (barefoot; averaged over four floor materials), QuietWalk reduces mean A-weighted noise level by 7.17 dB and peak noise level by 4.98 dB under a consistent recording setup. Cross-footwear experiments (barefoot, skate shoes, athletic sneakers, and high heels) across multiple surfaces further demonstrate robust adaptation to footwear-induced contact variations.

Problem

• humanoid locomotion
• ground reaction force
• footwear adaptation
• impact mitigation
• noise reduction

Innovation

• physics-informed reinforcement learning
• ground reaction force estimation
• inverse-dynamics-constrained PINN
• footwear-robust locomotion
• sensor-free impact mitigation