Indoor Occupancy Awareness and Localization Using Passive Electric Field Sensing

This paper introduces sparse low-power sensor networks using passive electric potential sensors for crowd-aware smart buildings. The proposed networks can detect human presence, track their positions, and estimate room occupancy by measuring the electric field disturbances caused by human movement. In addition, the proposed passive sensing system can remotely monitor human physiological signals of nearby stationary occupants (standoff ~1.5 m) when the ambient 60-Hz fields are filtered out. Based on an experimentally calibrated source range model, we further propose an improved multiple-occupant near-field indoor localization method using a maximum likelihood algorithm combined with dividing rectangular search. The tradeoff between search efficiency and localization accuracy is also discussed. Experimental results in two different rooms for 1 week duration show the effectiveness of our proposed sensing system: human identification with an accuracy of 98.3% on six participants, occupant localization methods with room occupancy estimation accuracy of 89.03%, and average localization error less than 0.3 m. These results open up new possibilities in security, indoor navigation or positioning systems, human-machine interfaces, and tele-health applications.