A linear-response and stretchable capacitive electronic skin integrated with machine learning for wireless health monitoring and smart robotic grasping

Yunkai Hu, Qijun Yu, Hao Xu, Haoran Gu, Shuo Zhang, Ke Shi, Jun Qian, Jun Li, Guangjie Yuan,
A linear-response and stretchable capacitive electronic skin integrated with machine learning for wireless health monitoring and smart robotic grasping,
Applied Materials Today,
Volume 47,
2025,
102970,
ISSN 2352-9407,
https://doi.org/10.1016/j.apmt.2025.102970.
(https://www.sciencedirect.com/science/article/pii/S2352940725003889)
Abstract: With the rapid development of wearable devices and embodied intelligence, the demand for the application of electronic skin (E-skin) is increasing in the fields of wireless health monitoring and smart robot grasping. On the basis of highly stretchable carbon nanotube (CNT) electrodes and a microporous thermoplastic polyurethane elastomer (TPU) dielectric layer, a stretch/pressure-sensitive (S/P-S) E-skin was fabricated with a linear response and gauge factor (GF) values of 0.63 and 1.3 × 10−2 kPa−1 in the strain and pressure ranges of 0–80 % and 0–51.6 kPa, respectively. Owing to its good performance, the E-skin was integrated into a wireless data glove and smart insole for the detection of the finger joint bending signal and foot pressure distribution, and machine learning-combined hand gesture and foot posture analysis systems were developed with high recognition accuracies of 98.6 % and 98.4 %, respectively. In addition, with the combination of a vision recognition module and a 3 × 3 pixel-pressure-sensitive (PP-S) E-skin, the smart robotic gripper successfully detected the gripping force and its distribution, and a machine learning-combined grasping and transfer stability analysis system was also developed with a high recognition accuracy of 98.5 %.
Keywords: Electronic skin; Linear-response; Machine learning; Wearable devices; Smart robotics