AI-Enabled Wireless Sensing for Temperature Monitoring of Cold Storage Facilities

Ensuring the integrity and safety of perishable goods within cold storage facilities has become a paramount concern for industries ranging from pharmaceuticals to food production. Meanwhile, as Integrated Sensing and Communication (ISaC) capable WiFi communication is on the horizon, We examine the feasibility of wireless sensing for cold storage monitoring. To this end, we aim to leverage the ubiquitous WiFi signals from commercial Internet-of-Things (IoT) devices, combined with lightweight Artificial Intelligence (AI), to monitor cold storage temperature without using dedicated IoT temperature sensors. Specifically, we build a sensing framework implemented on WiFi-enabled IoT hardware devices. The proposed framework uses the channel state information (CSI) of WiFi signals to monitor temperature variations within cold storage facilities by adopting a classification-based machine learning approach. Our experimental results empirically confirm the correlation between CSI and ambient temperature and evaluate the framework's performance at different locations within cold storage. Furthermore, the classification models have been successfully employed to accurately categorise temperature ranges with an accuracy of 92%, demonstrating the feasibility of wireless sensing techniques in monitoring cold storage facilities.