3D Bristle-Structured, Knitted-Fabric-Based Triboelectric Sensors for Machine Learning-Based Motion Recognition

With the development of electronic technology, triboelectric-based sensors have been widely researched in fields such as healthcare, rehabilitation training, and sports assistance due to their manufacturing convenience and self-powering advantages. Among them, 3D fabric-based triboelectric sensors not only possess advantages such as easy mechanized production, good breathability, and ease of wearing but also their unique 3D structure enhances the specific surface area, thereby amplifying the sensitivity. This study proposes a 3D bristle-structured fabric made by a digital knitting technology that has not been studied widely for triboelectric devices. By applying the 3D bristle structure with a large specific surface area to the single jersey fabric, the effective contact area during friction can be increased, resulting in a higher surface charge density. Additionally, the microcapacitor-like effect provided by the numerous microstructures allows the device to store more surface charge, further improving the output performance. The study systematically investigates the output performance of four different structures assembled by single jersey and 3D bristle-structured fabrics. The optimal sample exhibits a 57% higher output voltage than that of the reference 2D fabric sample. The 3D bristle-structured fabric demonstrates linear high sensitivity and distinct output performance when used as a sensor. Finally, a machine learning integration is applied to judge motion to assist a baseball pitcher in a self-training system.