Bioinspired High Linearity Fabric Sensors Enabled by Cross‐Scale Wrinkled Structure and Conductive Network Reconstruction for Wearable Electronics

Fabric‐based flexible strain sensors demonstrate significant potential for intelligent wearable devices, owing to their favorable breathability and wear comfort. However, achieving high linearity over a wide strain range remains challenging due to the stochastic evolution of microstructures or non‐uniform interfacial strain transmission in existing sensor designs. Inspired by the structure of an elephant's trunk, this study develops a novel MXene‐based strain sensor featuring a cross‐scale collaborative wrinkled structure. Benefiting from the synergistic effect of the dual‐scale wrinkles, the sensor exhibits outstanding linearity (0.9900) within a strain range of up to 110%, while also demonstrating high stretchability (>300%), high sensitivity (with a maximum gauge factor of 611.6036), an ultra‐low detection limit (0.1%), and remarkable durability (10 000 cycles). Moreover, with the assistance of deep learning algorithms, this sensor can recognize different gestures and finger bending angles with accuracies of 98.89% and 99.33%, respectively. To further verify its application potential, we have constructed an intelligent human‐machine interaction system that can achieve real‐time wireless control of the robotic hand and accurately distinguish different gestures, as well as recognize different objects and their size. The research shows that the proposed sensor has significant potential applications in motion monitoring and human‐machine interaction.