An AI ‐Based Wearable Computing System for Personalized Athletic Fatigue Analysis and Optimization via 6G Networks

With the rise of real‐time, data‐driven training in sports science, wearable computing systems have become essential for monitoring athlete performance and recovery. However, traditional systems often struggle with latency, scalability, and computational limitations, particularly in mobile and dynamic environments. While recent advancements highlight the potential of 6G networks, which offer ultra‐low latency and AI‐native edge computing, few studies have explored their integration into wearable systems for personalized fatigue monitoring. In this study, we propose a 6G‐enabled, lightweight wearable computing system that integrates a CNN‐BiLSTM architecture for personalized fatigue prediction. The model processes training load, sleep quality, and emotional state to predict fatigue, achieving a mean absolute error (MAE) of 1.0671 and a root mean square error (RMSE) of 1.3082, demonstrating its robust performance even with limited data. This system enables real‐time, on‐device inference for dynamic fatigue monitoring in athletic environments, providing scalable, low‐latency feedback that is critical for personalized performance optimization. Our work highlights the feasibility of 6G‐supported wearables for continuous fatigue tracking, offering a novel approach to personalized athletic training and rehabilitation.