Printed Strain Sensors for Motion Recognition: A Review of Materials, Fabrication Methods, and Machine Learning Algorithms

Abstract

Recent studies in functional nanomaterials with advanced macro, micro, and nano-scale structures have yielded substantial improvements in human-interfaced strain sensors for motion and gesture recognition. Furthermore, fundamental advances in nanomaterial printing have been developed and leveraged to translate these materials and mechanical innovations into practical applications. Significant progress in machine learning for human-interfaced strain sensing has unlocked numerous opportunities to improve lives and the human experience through healthcare innovations, sports performance monitoring, and human-machine interfaces. However, several key challenges still must be overcome if strain sensors are to become ubiquitous tools for human motion recognition. This review begins with a summary of the critical strain-sensing mechanisms employed today and how recent works have sought to push their boundaries. It then proceeds to cover the primary functional materials used in wearable strain sensors from a performance and printability perspective. Next is a review of recent advances in nanomaterial printing to produce the complex structures necessary for functional devices. Next, we summarize machine learning approaches for human gesture recognition and the myriad applications and use cases for human-interfaced strain sensors. Finally, it concludes with a discussion of challenges and opportunities for future research in the field.