Towards reliable intelligent occupancy detection for smart building applications

Occupancy detection is the core of smart lighting and smart ventilation systems with a significant role in energy cost reduction and improving occupants' comfort. Reliability and scalability are two main issues of the existing occupancy detection solutions. In this paper, we investigate the fusion of different sensing modalities to reduce false positive and false negative and hence improve the reliability. The proposed solution is being experimented for an office cubic but could be scaled up easily for the whole room. More specifically, we developed an alternative solution to passive infra-red (PIR) motion sensors so that it can only react to human presence rather than any moving object. We show that the combination of a distance sensor and CO2 could work as good as a PIR motion and also reduce the false detections when all sensors combined.