Semantic Discovery from Sensors and Image Data for Real-Time Spatio-Temporal Emergency Monitoring

Abstract

Modern district heating (DH) systems are essential contributors to large-scale city heating infrastructures. They consist of sensors, nodes, and methods for monitoring the status of the DH network. Sensing, processing, and analyzing data to locate an actual problematic emergency is a complicated task. This article presents the Spatio Temporal Emergency Monitoring (STEM) method that processes the semantic from multivariable sensors and infrared image data to locate real-time heating water leak emergencies in the DH network system. The sensory data includes multi-parameter DH network sensor data, such as water temperature, sweat rate, energy delivered, etc. The multimedia data is the infrared image data from a camera mounted on an unmanned aerial vehicle (UAV) for monitoring the environment at the underground DH network locations. The DH network monitoring system’s primary purpose is to automate the central heating network’s emergency monitoring. The presented real-time emergency monitoring approach processes the correlations between semantic and context of incoming data sets to determine an emergency’s actual risk of a water leak. The STEM approach processes the network’s sensors’ data and the corresponding multimedia image data into single observations to monitor real-time Spatio-temporal emergency events.