Triboelectric bending sensor based smart glove towards intuitive multi-dimensional human-machine interfaces

Highlights: • Development of high bending angle resolution triboelectric nanogenerator (BA-TENG) using PDMS and silicon rubber. • The signal-to-noise ratio (SNR) is improved by 19.36 dB with the assistance of a trans-impedance amplifier. • Multi-dimensional signal features are extracted, including valley (V), width (W), hold time (H), and time interval (T). • Light-brightness control, robotic hand control, and a virtual keyboard with user identification were achieved. • Machine learning technique (SVM) is used for user identification with an accuracy of 93.1%. Recent advances in human-machine interface (HMI) lead to a renewed interest in creating intuitive and immersive interaction. Here, we designed a simple-structured and high-resolution bending angle triboelectric sensor named bending-angle triboelectric nanogenerator (BA-TENG) to construct a glove-based multi-dimensional HMI. With the assistance of a customized print circuit board (PCB), the glove-based HMI exhibits high sensitivity and low crosstalk in real-time multi-channel finger motion sensing. The signal-to-noise ratio (SNR) is improved by 19.36 dB. By systematically extracting and analyzing the multi-dimensional signal features of the BA-TENG, intuitive multi-dimensional HMIs were realized for smart-home, advanced robotic control, and a virtual keyboard with user recognition functionality. The classification accuracy of the virtual keyboard for seven users reached 93.1% by leveraging the advanced machine learning technique. The proposed BA-TENG-based smart glove reveals its potential as a solution for minimalist-design and intuitive multi-dimensional HMI, promising in diversified areas, including the Internet of things (IoT), assistive technology, and intelligent recognition systems.