Fully-Printed Bionic Tactile E-Skin with Coupling Enhancement Effect to Recognize Object Assisted by Machine Learning

Abstract

In the pursuit of tactile sensation resembling human skin, the electronic skin (E-skin) has long been a subject of interest and inspires the exploration of various biomimetic structures. Nevertheless, the exceptional functionality of living organisms arises from the synergistic interplay of multiple internal factors, i.e. the coupling enhancement effect, which has received limited attention in existing studies. Here, a tactile E-skin featuring a multicoupled biomimetic structure that mimics three coupling elements found in the skin: Stratum spinosum, Meissner corpuscle, and Piezo2 protein, is proposed. By amalgamating their distinguishing characteristics, this bionic E-skin surpasses the performance of conventional counterparts such as sensitivity as high as 388.5 kPa−1, hysteresis as low as 0.76%, and response times as short as 10 ms. Furthermore, its fabrication methodology of efficient 3D printing shows great advantages in terms of production cost and customization. Finally, the sensor is expanded to a 9 × 9 pixels array for a machine learning-assisted intellisense system to recognize the fruits as a human does, achieving an accuracy of 91.4%. All of these prove the promising potential of this multicoupled biomimetic structure in wearable electronics, human–machine interface, soft robotics, and artificial sensing.