An Intelligent Braided Armband System for Training Assistance

Abstract With the increasing popularity of fitness training, correct posture and fatigue management are essential to prevent injuries and improve efficiency. Improper movements and excessive fatigue can cause muscle strain injuries and joint overload, increasing injury risk. However, traditional supervision relies on coaches’ subjective and intermittent assessments. Although sensing-based approaches have been explored, existing systems remain constrained by environmental sensitivity, motion occlusion, and insufficient long-term stability, restricting their use in fitness scenarios. Here, we report an intelligent braided armband system enabled by an elastic counter-pressure mechanism. This mechanism continuously provides elastic support to the arm while enhancing the sensing of muscle deformation and force variations. The braided composite structure directly transduces radial arm deformation into capacitance signals. Mechanical coupling amplifies signal responses, achieving high-sensitivity motion monitoring (8.17% kPa–1) and tensile interference resistance up to 50% strain. By integrating a hybrid learning architecture, the system enables data-driven signal decoupling, real-time motion recognition, and muscle fatigue estimation. The integration of structural design and algorithm development establishes a low-power and high-precision platform for fitness training and health monitoring.