FACET: Force-Adaptive Control via Impedance Reference Tracking for Legged Robots

2025-05-11

AI Summary

Legged robots exhibit rigid, unsafe responses during physical interaction due to the absence of force sensing, compromising safety and control precision. To address this, we propose an end-to-end reinforcement learning (RL) control framework based on impedance reference tracking—marking the first integration of learnable impedance reference trajectories into RL policy training, thereby unifying whole-body force interaction modeling with tunable compliance. Our method synergistically incorporates a virtual mass–spring–damper system, force-sensing closed-loop control, kinesthetic teaching, and physics-informed deployment optimization, enabling seamless transfer across quadrupedal and humanoid platforms. In simulation, the approach reduces collision impulse by 80% and maintains stability under impulsive disturbances up to 200 N·s. Physical experiments demonstrate compliant dragging of loads up to two-thirds the robot’s body weight and safe, robust human–robot interaction.

Technology Category

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

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
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Reinforcement learning (RL) has made significant strides in legged robot control, enabling locomotion across diverse terrains and complex loco-manipulation capabilities. However, the commonly used position or velocity tracking-based objectives are agnostic to forces experienced by the robot, leading to stiff and potentially dangerous behaviors and poor control during forceful interactions. To address this limitation, we present emph{Force-Adaptive Control via Impedance Reference Tracking} (FACET). Inspired by impedance control, we use RL to train a control policy to imitate a virtual mass-spring-damper system, allowing fine-grained control under external forces by manipulating the virtual spring. In simulation, we demonstrate that our quadruped robot achieves improved robustness to large impulses (up to 200 Ns) and exhibits controllable compliance, achieving an 80% reduction in collision impulse. The policy is deployed to a physical robot to showcase both compliance and the ability to engage with large forces by kinesthetic control and pulling payloads up to 2/3 of its weight. Further extension to a legged loco-manipulator and a humanoid shows the applicability of our method to more complex settings to enable whole-body compliance control. Project Website: https://egalahad.github.io/facet/

Problem

• Addresses stiff robot behaviors from force-agnostic control objectives
• Enables fine-grained force adaptation via virtual impedance tracking
• Improves robustness to large impulses and collision forces

Innovation

• RL-trained policy imitates virtual mass-spring-damper system
• Fine-grained force control via virtual spring manipulation
• Achieves 80% collision impulse reduction in simulation