Multimodal flexible sensor system: long term stable monitoring and multi-tasking system using a machine learning

Abstract

This study aimed to develop wireless multi-modal and multi-tasking flexible sensor sheets. To realize multimodal flexible sensor, various sensors using low-cost solution-based process were proposed for integration on flexible films. In particular, monitoring of continuous vital signals was demonstrated by attaching a sensor sheet to a human. For multitasking systems to make X-in-1 sensors, reservoir computing, which is one of the machine learning techniques, was applied to a flexible sensor. As a proof-of-concept, weather sensors that detect the rain droplet volume and wind velocity were developed. Although there are remaining issues that must be resolved to realize practical devices, the proposed method can be used to build a multimodal and multitasking low-power sensor system in the future.