Abstract

A PAM/AG/TR/CNTs composite hydrogel was fabricated for strain and pressure sensors. • The addition of trehalose enhances frost resistance of the composite hydrogel. • The hydrogel demonstrates flexibility, strong adhesion, and good biocompatibility. • The hydrogel sensors are capable of monitoring diverse physiological signals. Hydrogel sensors hold great potential for flexible wearable devices. However, their practical application in complex environments is limited by several shortcomings, such as the inability to detect tensile and compressive stimuli with high accuracy, insufficient frost resistance, and a lack of self-adhesiveness. Herein, we incorporated trehalose into a dual-network hydrogel system composed of polyacrylamide, agarose, and carbon nanotubes through a simple one-pot method to produce a novel composite hydrogel. The addition of trehalose significantly enhances both the mechanical properties and frost resistance of the composite hydrogel to deliver ultrahigh flexibility, superior stretchability and compressibility, strong adhesion, excellent biocompatibility, and reliable frost resistance. Sensors assembled from the composite hydrogel provide high sensitivity with antifreeze performance to temperatures as low as −30 °C. The optimized strain sensor delivers a gauge factor of 9.84 within the 60–90% strain range, a low detection limit (0.1%), and a broad sensing range (0.1–90%). The optimized pressure sensor exhibits high sensitivity (3.46 kPa −1 in the 1–3 kPa region), a wide detection range (39.4 Pa to 100 kPa), fast response (70 ms), and reproducible performance over 1000 cycles. In practical applications, the sensors successfully detect diverse physiological signals, ranging from arterial pulse pressure to heavy foot-treading pressure. Further integration with the Multi-Layer Perceptron (MLP) machine-learning algorithm results in exceptional prediction accuracy of gesture recognition, achieving 99.44%. All these features testify to promising pr