Occupancy Determination by Backscattered Visible Light Sensing

Demand driven control of heating or cooling, has the potential to reduce the energy consumption of buildings to a significant portion. One of the main cornerstones of such a demand driven control is the accurate determination of occupancy in the building or subareas of the building. In recent years many different approaches, founded on various technologies and sensing principles have been used to establish an occupancy detection system. Especially in existing buildings, retrofitting such a system can be associated with high installation effort and cost. Furthermore, some technological approaches, mainly camera-based systems are inflicting concerns in regard to the privacy of the users. In this manuscript, we present and discuss an approach for a lighting fixture that performs person recognition and occupancy determination based on the technology of Visible Light Sensing. In our implemented system by solely capturing reflections of the visible light, caused by bypassing persons, the occupancy, the movement direction and the walking speed can be accurately determined. Based on comprehensive real world experiments we will show how our system, fully integrated into an off-the shelf luminaire, is able to operate in different scenarios and ambient light conditions and only requires minimal installation effort without causing any privacy concerns.