Wearable devices that can accurately detect the position and strength of applied pressure have been a focus of attention in the field of flexible tactile sensing. However, conventional sensor arrays with multiple units have many challenges in practical applications due to their complex circuit layouts and poor robustness, etc., which could be solved by minimizing the number of connecting electrodes among each unit. Herein, a pressure sensor with geometrically asymmetric structure has been proposed to remove all the connected inner electrodes. The sensor is constructed by three pieces of fabrics, with one insulating fabric sandwiched by two conductive fabrics coated with carbon nanotubes. Notably, the bottom fabric is specially designed in an optimized asymmetric shape with a rectangular hole located off-center of the square, assisting the sensor to generate different resistive and capacitive responses among pre-divided nine areas through only three external electrodes. Compared to conventional resistive sensor arrays, the number of electrodes is reduced by 83.3 %. With the assistance of machine learning, pressure-positioning with an accuracy of more than 98.1 % and precise strength measurements are achieved. Based on this sensor and the trained neural network, a two-level password system is constructed for more reliable role recognition. The asymmetric strategy proposed in this work provides a novel design for realizing precise detection of both position and strength using extremely simplified circuits, which has great potential in the field of wearable electronics.
Read full abstract