Toward intelligent wearables: engineering self-powered triboelectric nanofiber sensors for the classification of...

Triboelectric nanogenerators (TENGs) have emerged as promising platforms for self-powered sensing and real-time biomechanical monitoring. However, current ankle injury classification systems lack both machine learning intelligence and self-powered operation, limiting their effectiveness in dynamic environments. Here, we report a dual-filler strategy in electrospun PVDF-HFP nanofibers, incorporating copper sulfate and graphite to enhance the surface contact points by reducing fiber diameter, while improving dielectric polarization and stability This design yields a six-fold performance improvement compared to pristine fibers (50 V), delivering outputs of ∼302 V, 9.1 µA, and 80.6 nC, with excellent durability over 10 000 cycles, and a peak power density of ∼1.28 W m^-2 sufficient to charge capacitors and power small electronics. Integrated into an ankle-worn platform, the fiber-based TENG device generated high-fidelity biomechanical signals. When analyzed using machine learning algorithms, the system achieves up to 99% accuracy across 700 datasets in detecting risky motions preceding sprains. This intelligence shifts the system from passive monitoring to proactive prevention, providing actionable feedback before injury onset. Beyond ankle injuries, this convergence of self-powered materials and artificial intelligence establishes a new class of intelligent wearables and paves the way for advanced musculoskeletal health monitoring and preventive medicine.