Inferring Occupants Count from Low Power IP based Devices in Building through Data Fusion
Dipti Trivedi, Venkataramana Badarla · 2019
Building automation and building climate control efficiency can be improved by applying accurate occupancy information into the building management system. Extracting the occupancy status from the systems already existing in the building has a great potential to provide precise information with lower cost. The present work proposes an occupants count inference system using low power IP based devices carried by the occupants. Various infrastructure-based sensing elements are used to detect the IP based devices carried by occupants in the building environment. Here, we propose a mass assignment method which is based on the probabilistic approach and relative trust in each sensing element to infer the occupants' count in the zone. This paper presents some preliminary results, with experimental verification, in this direction.