QuietPaw: Learning Quadrupedal Locomotion with Versatile Noise Preference Alignment

2025-03-06

AI Summary

High-mobility locomotion of quadrupedal robots generates substantial footstep noise, hindering deployment in human-dense environments such as homes and hospitals. Method: We propose the Conditioned Noise-Constrained Policy (CNCP), a novel framework enabling continuous, simultaneous trade-off control between acoustic noise and agility within a single policy. CNCP employs a noise-level-conditioned policy architecture; value function representation decomposition decouples state and noise-condition embeddings, enabling zero-shot generalization across noise levels without retraining. Constraint-based reinforcement learning is used to optimize the Pareto frontier of noise-agility performance. Results: Evaluated in simulation and on real quadruped platforms, CNCP achieves 15–22 dB(A) reduction in footstep sound pressure level while retaining over 90% of baseline locomotion performance—significantly enhancing human-robot coexistence capability.

Technology Category

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

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Agentic search
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems

Abstract

When operating at their full capacity, quadrupedal robots can produce loud footstep noise, which can be disruptive in human-centered environments like homes, offices, and hospitals. As a result, balancing locomotion performance with noise constraints is crucial for the successful real-world deployment of quadrupedal robots. However, achieving adaptive noise control is challenging due to (a) the trade-off between agility and noise minimization, (b) the need for generalization across diverse deployment conditions, and (c) the difficulty of effectively adjusting policies based on noise requirements. We propose QuietPaw, a framework incorporating our Conditional Noise-Constrained Policy (CNCP), a constrained learning-based algorithm that enables flexible, noise-aware locomotion by conditioning policy behavior on noise-reduction levels. We leverage value representation decomposition in the critics, disentangling state representations from condition-dependent representations and this allows a single versatile policy to generalize across noise levels without retraining while improving the Pareto trade-off between agility and noise reduction. We validate our approach in simulation and the real world, demonstrating that CNCP can effectively balance locomotion performance and noise constraints, achieving continuously adjustable noise reduction.

Problem

• Balancing locomotion performance with noise constraints in quadrupedal robots.
• Achieving adaptive noise control across diverse deployment conditions.
• Improving trade-off between agility and noise reduction in robots.

Innovation

• Conditional Noise-Constrained Policy (CNCP)
• Value representation decomposition in critics
• Single versatile policy for noise generalization