🤖 AI Summary
Traditional personalized exercise planning relies heavily on expert-driven iterative refinement, incurring high costs and limited accessibility. To address this, we propose a large language model (LLM)-based closed-loop conversational agent that dynamically models users’ fitness goals, time constraints, and health limitations via natural language interaction. The agent integrates domain-specific sports medicine knowledge and guideline alignment mechanisms to autonomously generate interpretable, iteratively refinable weekly training plans compliant with the American College of Sports Medicine (ACSM) evidence-based standards. Our key contributions include: (i) the first closed-loop conversational paradigm explicitly designed for exercise planning; and (ii) a multi-dimensional evaluation framework combining user studies, intrinsic metrics, and expert review with dialogue state tracking. Experimental results show that 92% of users successfully received high-quality plans; 87% of generated plans were validated by sports medicine experts as ACSM-compliant; and the plans significantly outperformed baselines in executability and individual adaptability.
📝 Abstract
Creating personalized and actionable exercise plans often requires iteration with experts, which can be costly and inaccessible to many individuals. This work explores the capabilities of Large Language Models (LLMs) in addressing these challenges. We present PlanFitting, an LLM-driven conversational agent that assists users in creating and refining personalized weekly exercise plans. By engaging users in free-form conversations, PlanFitting helps elicit users' goals, availabilities, and potential obstacles, and enables individuals to generate personalized exercise plans aligned with established exercise guidelines. Our study -- involving a user study, intrinsic evaluation, and expert evaluation -- demonstrated PlanFitting's ability to guide users to create tailored, actionable, and evidence-based plans. We discuss future design opportunities for LLM-driven conversational agents to create plans that better comply with exercise principles and accommodate personal constraints.
