Multimodal Perception and Machine Learning-Empowered Human Machine Interfaces With Double-Network Hydrogel Fibers

With the rapid development of intelligent technologies, human-machine interfaces (HMIs) are increasingly integrated into modern society, enabling efficient communication between humans and robots and thus building intelligent living. Flexible sensors are key signal conversion materials in these systems. However, materials that simultaneously possess high tensile strain, sufficient toughness, and fiber morphology compatibility suitable for wearable devices remain scarce. Therefore, this work develops a kind of double-network polyacrylamide-alginate (PAM-Alg) hydrogel fiber sensors, which combines high elasticity (∼700% strain), enhanced tensile strength, and inherent softness and wearability. The shell-less fiber structure achieves a uniform stress distribution and is easily woven into textile wearables, thus endowing the sensor with excellent mechanical strength and long-term sensing stability. These hydrogel fibers, combined with a dexterous robotic hand and a one-dimensional convolutional neural network, can accurately identify objects from multimodal grasping signals and achieve reliable temperature differentiation. Furthermore, the hydrogel fibers can also be used as a flexible electrode for triboelectric nanogenerators to achieve energy harvesting and material recognition. This work establishes a scalable fiber-based hydrogel platform for intelligent wearable systems, soft robotics, and next-generation human-machine interaction technologies.