As the aging population increases in many countries, electronic skin (e-skin) for health monitoring has been attracting much attention. However, to realize the industrialization of e-skin, two factors must be optimized. The first is to achieve high comfort, which can significantly improve the user experience. The second is to make the e-skin intelligent, so it can detect and analyze physiological signals at the same time. In this article, intelligent and multifunctional e-skin consisting of laser-scribed graphene and polyurethane (PU) nanomesh is realized with high comfort. The e-skin can be used as a strain sensor with large measurement range (>60%), good sensitivity (GF≈40), high linearity range (60%), and excellent stability (>1000 cycles). By analyzing the morphology of e-skin, a parallel networks model is proposed to express the mechanism of the strain sensor. In addition, laser scribing is also applied to etch the insulating PU, which greatly decreases the impedance in detecting electrophysiology signals. Finally, the e-skin is applied to monitor the electrocardiogram, electroencephalogram (EEG), and electrooculogram signals. A time- and frequency-domain concatenated convolution neural network is built to analyze the EEG signal detected using the e-skin on the forehead and classify the attention level of testers.
Keywords: crack models; laser scribing graphene; nanomeshes; neural networks; physilogical signal monitoring.
© 2021 Wiley-VCH GmbH.