Calgary

University's new AI security system that studies movement a 1st for Canada

A new artificial intelligence security system is being installed for the first time in Canada at a Calgary university.

'When that pattern breaks, what it does, that screen comes to life'

A security officer at Mount Royal University scans the school's new AI security camera system. (Dave Gilson/CBC)

A new artificial intelligence security system is being installed for the first time in Canada at a Calgary university.

The network of high-resolution cameras paired with AI, also known as black screen monitoring, will spend two weeks learning normal movement patterns so it can scan for anomalies on campus — like the hurried patterns of students once they spot a person nearby carrying a gun.

"Now when that pattern breaks, what it does, that screen comes to life and it shows the people in the security office where the pattern is now different and then it's up to a human being to decide what to do about it," says Grant Sommerfeld, associate vice-president of facilities management at Mount Royal University. 

The 360-degree cameras will keep an eye out for anomalies 24/7 across campus. (Dave Gilson/CBC)

That's where the black screen name comes from — the black screens only come to life when something odd is detected.

The iCetana system will learn patterns based on what's normal for different locations at different times of day, and on the weekdays and weekends. And the 360-degree cameras make it impossible to enter the campus without coming into view of at least one of the cameras.

The university's security director, Peter Davison, said studies show a single security officer can really only monitor two camera feeds at once.

"Now, the dispatcher sitting at the console is no longer looking at 300 images at a time. This has allowed them to concentrate on just the certain things that pop up," he said.

The black screen monitoring system only wakes when it notices something unusual, to flag to a human employee. (Dave Gilson/CBC)

It will gather data like people counts to help map foot-traffic patterns, and has thermal imaging to detect heat before it becomes a fire.

So far, the system has flagged a man who fell and couldn't call for help, a vehicle doing donuts, and two men play-fighting, according to the university's website.

'We don't believe there are privacy concerns': school

Sommerfeld says the system doesn't monitor individual people and that it adheres to privacy regulations.

"It doesn't look at men or women, age groups or anything. It just looks at these pixels and in that regard, we don't believe there are privacy concerns with it."

But, a person's movement can be tracked across campus — for example, if someone is seen stealing a bike, their path can be followed through the camera system.

That raises some red flags for MRU student Aria Burrell.

"Without more details given to students as to how the artificial intelligence will be identifying them and what criteria will be used by security, I'm concerned about that," Burrell said.

With files from Dave Gilson