Multimodal Flexible Sensor for the Detection of Pressing–Bending–Twisting Mechanical Deformations
Chen Yang, Hui Liu, Jin Ma, Ming Xu
Abstract
Flexible sensors are increasingly significant in applications such as smart wearables and human-computer interactions. However, typical flexible sensors are spatially limited and can generally detect only one deformation mode. This study presents a novel multimodal flexible sensor that combines three sensing units: optoelectronics, ionic liquids, and conductive fabrics. It employs a sophisticated superposition and combination of the three sensing methods to achieve up to eight mechanical deformations, including pressing, bending, twisting, and combinations thereof, all within a very small sensor space. This sensor has excellent detection performance, high sensitivity (optoelectronics 4.312, ionic liquid 8.186, conductive fabric 2.438), a wide measurement range (pressing 0-75 kPa, bending 0-90°, and twisting 0-180°), and good consistency and repeatability. To address the signal coupling problem in multimode sensors, a deep learning method based on the Transformer is combined to provide precise decoupling of multimode signals and high-precision characterization of each mechanical deformation. Finally, the wrist joint experiments demonstrate the sensor's versatile uses in human-computer interaction.