Policy Learning for Social Robot-Led Physiotherapy

2025-09-14

AI Summary

In physical therapy, scarce real-world patient behavioral data hinder social robots from delivering personalized, autonomous guidance. Method: We propose a simulation-driven reinforcement learning framework that employs a clinician-validated patient behavior agent model to generate high-fidelity synthetic data. This model integrates exertion tolerance estimation and subjective fatigue scoring prediction, enabling policy training in simulation for generalization across multiple rehabilitation stages. Contribution/Results: Our approach is the first to explicitly embed clinical expertise into patient behavior modeling, eliminating reliance on limited real-world data. It enables real-time, dynamic adjustment of exercise instruction intensity. Experiments demonstrate that the learned policy significantly improves guidance adaptability and training safety across diverse simulated patients and rehabilitation stages. The framework establishes a scalable, clinically deployable foundation for autonomous decision-making in robotic-assisted rehabilitation.

Technology Category

• Intelligent Robots: Behavior Learning & Control
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Cognitive Modeling & Cognitive Systems: Simulating Human Behavior

Application Category

• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Social robots offer a promising solution for autonomously guiding patients through physiotherapy exercise sessions, but effective deployment requires advanced decision-making to adapt to patient needs. A key challenge is the scarcity of patient behavior data for developing robust policies. To address this, we engaged 33 expert healthcare practitioners as patient proxies, using their interactions with our robot to inform a patient behavior model capable of generating exercise performance metrics and subjective scores on perceived exertion. We trained a reinforcement learning-based policy in simulation, demonstrating that it can adapt exercise instructions to individual exertion tolerances and fluctuating performance, while also being applicable to patients at different recovery stages with varying exercise plans.

Problem

• Adapting robot-led physiotherapy to patient needs
• Addressing scarcity of patient behavior data
• Developing personalized exercise instruction policies

Innovation

• Reinforcement learning policy trained in simulation
• Expert healthcare practitioners as patient proxies
• Adapting instructions to exertion and performance