A multifunctional, double network cellulose-acrylamide based hydrogel sensor reinforced using a liquid metal for human motion detection and dual-mode handwriting recognition via a transfer learning algorithm

Abstract

Conductive hydrogels have received extensive attention as a candidate material for flexible strain sensors. Especially, cellulose hydrogels have good flexibility, biocompatibility, and environmental degradability, and therefore can be widely applied in wearable electronics, human–machine interfaces, and soft robotics. However, the typical single network cellulose hydrogel sensor reveals low mechanical loading, greatly limiting practical applications. It is still a great challenge to develop a cellulose conductive hydrogel sensor with excellent mechanical properties and ideal conductivity. Herein, a novel multifunctional conductive hydrogel based on cellulose, acrylamide and liquid metal (LM) was prepared via a one-pot synthesis strategy. Acrylamide monomers were combined with N,N′-methylenebisacrylamide (MBA) to form a robust three-dimensional polyacrylamide (PAM) network. The acrylamide was then physically cross-linked with cellulose. LMs as conductive fillers embedded in the network of the hydrogel can not only enhance the stability of the network structure, but also improve conductivity. This resulted in high tensile properties (963%), superior strain responses (with a gauge factor of 21.7), a fast response time of 240 ms and a 90% retention of performance after 600 cycles. Moreover, the as-prepared hydrogel sensor can continuously monitor human motion and measure the contraction of muscles in different parts of the body including biceps, triceps and pectoralis major. By integrating the transfer learning algorithm (Resnet50), an intelligent dual-mode handwriting recognition system was developed for detecting finger touch signals (compressive-contact mode) and finger wiggle signals (finger-strain mode), with both high accuracy (100% for ten numbers and 98% for 26 letters) and fast recognition time (<1 s) when recognizing Arabic numbers. This study showed that the multifunctional cellulose hydrogels will have broad application prospects in developing intelligent human–machine interfaces, virtual reality interactions, and future bioelectronics.