Robust detection of persons in emergency situations in public buildings
Jan Haase · 2021
This paper introduces an automated robust indoor detection system that is able to detect persons in emergency situations. Other than comparable projects, occupancy levels are not only measured by data fusion and bias correction algorithms, but also extended with sensor credibility. Based on trustworthiness of sensor nodes, a metaheuristic estimates plausible occupancy estimations of the building. According on whether sensor telemetry is valid and plausible, trust is degraded or raised, letting the system react to defect devices and occurring calamities. In addition, its sensor network is designed openly, making it easy to integrate new, previously unknown, sensors to the mesh. A sample web-based frontend is presented, which can be used to easily view, simulate and manipulate the algorithm. To validate the result, disaster scenarios were tested in a simulated public building environment.