EdgeThermal

EdgeThermal is an Edge AI app that runs thermal-image anomaly detection directly on-site, turning existing thermal cameras into early-warning systems for overheating motors, electrical panels, bearings, and conveyor components. It ingests RTSP/ONVIF streams, learns a baseline “normal heat map” per asset, then flags deviations with simple, explainable overlays and thresholds operators can trust. Processing happens on an edge box (Jetson/Intel NUC) so plants avoid cloud latency, bandwidth costs, and data-sharing headaches. The app includes a lightweight dashboard for setup, zone masking, alert routing, and maintenance notes. This is not a magic AI that predicts every failure; it focuses on a narrow, high-value problem: catching abnormal heat patterns early enough to prevent downtime and safety incidents. It’s designed for small-to-mid industrial sites that can’t afford complex enterprise platforms.