The National Renewable Energy Laboratory in Golden, Colo., has come up with new technology to eliminate a problem with occupancy sensors: the sensors sometimes mistakenly turn out the lights when people are working.

For 30 years, occupancy sensors have relied primarily on motion detection, but if there isn’t a lot of movement, the sensors don’t detect people, creating a nuisance.

NREL has now developed and made available for license the Image Processing Occupancy Sensor (IPOS), which combines an inexpensive camera and computer vision algorithms that can recognize the presence of human occupants, according to NREL, which says IPOS raises the accuracy of occupancy detection from about 75 percent to the upper 90 percent range. The sensors will likely sell for between $100 and $200 once a licensee starts producing them in volume.

IPOS can detect with almost 100 percent accuracy the number of people in an area, spots where there are no people and the level of illumination. IPOS communicates its information with building automation systems via standard protocols. Whether the sensors are installed in an office or a big-box store, they can detect precisely where people are located, leaving the lights bright in those areas and dimming the lights in unoccupied spaces. This can make a significant dent in energy bills for facilities with large square footage.

The information from IPOS can also be used to make decisions on the amount of ventilation, air conditioning, or daylighting needed, and it can signal when security officers should be alerted.

IPOS is based on commercially available embedded hardware widely used by the smartphone industry. IPOS’s field of view is a 45-degree angle, similar to most cameras, and its range is up to 100 feet, much longer than traditional occupancy detectors with a 20-foot range. IPOS can replace several traditional occupancy sensors by segmenting its images into up to 16 virtual zones.

To avoid privacy complaints, the camera-based technology captures the image, analyzes it, and then destroys it soon after processing, so the image never leaves the device.

IPOS is currently being tested by NREL in a few environments, including a large retailer in Centennial, Colorado, that serves as a test bed for innovations that could later be launched nationwide.