WiFi Sensing Based Textile Wetness Monitoring

The Internet-of-things (IoT) for environmental sensing has attracted substantial interest and has great potential in many applications for both industry and lifestyle. Our research aims to sense the wetness of textiles via wireless transmission, which could provide a non-invasive alternative to the existing wearable sensing techniques. We proposed a sensing framework that uses the channel state information (CSI) signature of ambient WiFi signals to reflect the wetness of the targeted textile with a classification-based machine learning approach. As a proof of concept, we experimentally implemented the proposed frame-work with commodity hardware to affirm the wide deployability. Our embedded system-based design empirically demonstrated the proposed WiFi textile wetness sensing was feasible and could predict the wetness levels of Oml to 20ml with a resolution of 5ml for the target textile with more than 99% accuracy for indoor environments. We tested the robustness and scalability of our wireless wetness sensing technology via this real-world implementation and achieved over 90% accuracy with up to a 6m working range in an open outdoor environment. Overall, our prototype implementation provides an affordable, non-invasive, and device-free solution that is perfect for many household healthcare applications, such as monitoring sweat conditions and detecting body fluid leakage for infants and elderly care.