Duke University depends on thousands on surveillance cameras, which it relies on to be the university’s eyes. The cameras watch over its campus, students and employees, so that if an incident occurs, Duke will be armed with footage of what happened.
However, this technology can be imperfect. It is possible for these cameras to be tilted, blurred or obstructed by a glare at any moment, thus impairing the function of these vital security tools.
Theoretically at least, a simple answer to this problem would just be to check every camera every day. But for an institution with hundreds or even thousands of cameras, this task is laborious and subject to human error. That’s where Durham-based Ai-RGUS comes in.
Named after the all-seeing giant from Greek mythology, Argus, Ai-RGUS is an artificial intelligence-based software that monitors security cameras and makes sure that they are all working according to a base image. The “i” in Ai-RGUS is silent, and it indicates that the technology uses AI.
It was designed by CEO Daniël Reichman while he was completing his Ph.D. in electrical and electronics engineering at Duke in order to help the university keep its then-1,100 cameras in working order.
“It’s a huge investment to put up 1,100 cameras,” Reichman said, “so they want to make sure that once they have them, they can actually count on having the images from the cameras.”
Ai-RGUS keeps security cameras in check by first requiring users to take a base image that the software will use as a comparison to the current image it sees. Potential problems that Ai-RGUS could identify in an image are blur, tilt, tamper, obstruction, glare and a blank/black screen.
If Ai-RGUS identifies a problem in a camera it will mark it as “unwell,” so its owner can fix it. Ai-RGUS also keeps track of when cameras are habitually “unwell” to help owners to identify faulty cameras for potential replacement.
Ai-RGUS is used by all kinds of institutions that use a large number of security cameras, from college campuses—like Duke and the University of Pennsylvania—to stores like Nevada-based Green Valley Grocery’s 50+ locations, and even the City of Las Vegas. It uses a subscription model with an installation fee that varies with the scope, size and type of each camera system.
Over the past year, Ai-RGUS has expanded its team into Las Vegas and was a semifinalist for an NC IDEA SEED grant. While Ai-RGUS originally sold its software directly to businesses and institutions, it now also partners with camera installation services, which can include Ai-RGUS at the time of installation as a bonus service.
“It’s the kind of thing where you can set it and forget it,” Reichman said. “It sends you emails when it needs to. Otherwise, don’t bother looking at it. This is such a complicated and fractured space, and we’re trying to make it as simple as possible. It was such a headache to install the cameras. We don’t want to add more burden; it’s quite the opposite.”