AI-Enhanced Flexible Tactile Sensors with Multidimensional Perception

Electronic skin possesses unique advantages such as high sensitivity, ultra-thinness, flexibility, and conformal attachment. When integrated with artificial intelligence (AI) technology, its value and applications are further enhanced. However, current research faces several limitations, such as single sensing capability, poor stability, and difficulties in integration with AI. To address these issues, this paper presents a tactile sensor based on silver nanowires (AgNWs) suspended strain membranes and micropillar support structures, which demonstrates excellent pressure sensing capabilities ($>4.17 \Omega \cdot \text{kPa}^{-1}$) and strain sensing capabilities ($\text{GF}>95$) while maintaining stable performance over 10,000 cycles. Leveraging these advanced features alongside seamless AI integration, our sensor achieves sub-millimeter-level reconstruction of mahjong tile surfaces and highly precise texture identification, with an accuracy exceeding 95%. This establishes it as a promising platform for next-generation intelligent tactile systems with vast application potential.