Smart Occupancy Sensor Debuts

Smart Occupancy Sensor Debuts

NREL's R&D 100 Award-winning occupancy sensor offers a low-cost, high-impact solution to a pressing market challenge: improving building energy performance.

A photo of the Image Processing Occupancy Sensor, a white plastic device similar in size and shape to a smartphone. On the top left of the device is a small camera lens. On the top right side of the device is a three-inch black antenna. The exterior casing of the device is opaque, exposing the integrated circuit (IC), or microchip, within. A cord and a USB charging port are plugged into an area on the side of the device where one edge of the IC is exposed. In the background is a blackboard on which computer algorithms are written. Enlarge image

NREL's innovative IPOS technology combines a low-cost camera and computer vision algorithms to boost the accuracy of building occupancy detection by more than 20% for improved building energy performance.
Photo by Dennis Schroeder, NREL

Noted as one of the 100 most significant innovations of 2013 by R&D Magazine, the Image Processing Occupancy Sensor (IPOS) developed at NREL provides a groundbreaking solution to the challenge of improving building energy performance.

Commercial buildings and industrial plants account for roughly 50% of U.S. energy consumption, which translates to roughly $400 billion in energy costs—and a substantial portion of the carbon emissions that contribute to climate change. According to the U.S. Department of Energy (DOE), meeting President Obama's Better Buildings Challenge goal of reducing the energy intensity of all U.S. commercial buildings and industrial facilities just 2.5% annually would result in an annual savings of more than $80 billion after a decade while significantly reducing the collective carbon footprint of those facilities.

Boosting Occupancy Sensor Effectiveness

Occupancy sensors are one tactic for cutting building energy waste, but until now their effectiveness has been limited. With its novel approach to occupancy detection, the IPOS boosts accuracy more than 20% while providing a richer set of information for increasing building energy efficiency. That represents billions in potential energy cost savings.

For 30 years, occupancy sensors relied primarily on passive infrared (PIR) technology to infer the presence of people by detecting motion. This limited their range and accuracy, explained Senior Engineer Luigi Gentile Polese, one of the three-man team at NREL that developed the IPOS. False positives and negatives caused annoyances, such as lights shutting off when employees were working late. This often led people to disable the sensors, negating their energy saving benefits.

To address these limitations, NREL tapped its Commercial Buildings Research Group's deep expertise in building energy efficiency and software development to create an innovative approach to occupancy detection. Polese, along with Larry Brackney and Alex Swindler, began exploring the use of computer vision in occupancy sensors. Leveraging the low-cost camera and microprocessing technologies used in smartphones, they combined image processing capabilities with computer software to develop the IPOS. Initially funded by DOE, the team later partnered with Bonneville Power Administration to make the solution even more robust.

Proverbial Eye in the Sky

An image of an infrared motion sensor screenshot capturing motion activity in blue with three red circle and one white circle indicating activity. Enlarge image

Unlike conventional passive infrared motion sensors used to infer occupancy, IPOS uses sequential image subtractions like this one for extracting and analyzing motion-dependent occupancy content.
Image from NREL

Relying on digital images, the IPOS can detect occupancy with 99% accuracy across a much wider area than PIR sensors, Polese said. Furthermore, it can segment that area into as many as 16 virtual zones and identify the number of people, the level of activity, and the amount of light in each. All of this allows for highly targeted control of lighting, heating, and other building systems based on occupant need.

"The ability to extract a rich set of feedback from a single device creates unprecedented potential to cut energy costs," said Ty Ferretti, a licensing executive in NREL's Commercialization and Technology Transfer Office. As a result, he said, the IPOS has captured the attention of some major industry players and brought a number of potential licensees to the table.

The level of interest has been "through the roof," said Ferretti, who nominated the IPOS for the R&D 100 Award. "The market is just itching for it. Everybody needs to save money, and in any business, energy efficiency is the low-hanging fruit." And the IPOS offers big bang for the buck, he said. "This is a simple, low-cost technology that can replace existing sensors and deliver substantial savings with added functionality."

Cutting Energy Use, Costs

The inherent flexibility of the IPOS creates ample opportunities for customization, Ferretti said. One example is the ability to integrate the software into existing building security systems, further increasing potential capital savings. Other applications under consideration include elevator operations, medical sterilization, retail video displays, energy performance metering, and more.

In each case, the IPOS offers a low-cost, high-impact way to cut building energy waste with a relatively low up-front capital investment. And the return on that investment extends beyond corporate balance sheets to the larger U.S. goal of reducing carbon pollution in support of the President's Climate Action Plan.

—Written by Karen Petersen

Top of page