NSF Award
1448573
This Small Business Innovation Research phase I project will investigate a wireless human presence detection (WHPD) system that makes use of extant metadata to enable devices - in the near term, vents controlled by short-range wireless networks - to program themselves and respond to changes in users' habits and environment without the addition of any new hardware. A vent that can build a predictive schedule can close when a user is not likely to be present. For example, people rarely use their kitchen or den when sleeping at night. Since closing these vents intelligently has been shown to save between 14% and 30% of a heating, ventilation, and air conditioning (HVAC) system's run-time, a way to automatically close them will save money, energy, and improve user comfort. Per the 2009 U.S. Census, there are 97 million American homes with a central HVAC system and 61% of all American households have a home Wi-Fi network. With an average of four rooms over-conditioned at various times, there are at least 160 million rooms at suboptimal temperatures. Based on these figures, and a sub $100 retail price for such a device, the U.S. market size is roughly $13 billion. <br/><br/>The intellectual merit of this project will hinge on the capability to turn extant wireless metadata into effective wireless human presence detection (WHPD) as an enabling technology capable of enhancing many different technologies already used in home/commercial/industrial automation. Current presence-detecting methodologies usually require motion detection to function. There are presence detectors that identify people by their thermal, or visual light shapes (object recognition), but these devices are unfit for consumer applications due to cost and complexity. This project will result in an algorithm to detect presence reliably, or semi-reliably, from received signal strength indicator (RSSI) or any other method that does not require specialized hardware. The broader impact/commercial potential of a wireless presence detector enables many more technologies as well. For instance, the Philips Hue light bulb is a device on its own ZigBee powered network. When integrated with the proposed WHPD system, the Hue could function as a motion sensing light, turning itself off when no one is in the room. Similarly, any connected thermostat could become a learning thermostat, using the WHPD data supplied by this system as a way to improve home heating and cooling efficiency.
