The software securing surveillance

International Security Journal sits down with Daniël Reichman, CEO and Chief Scientist, Ai-RGUS.

Can you tell us about the background behind Ai-RGUS?

I was wrapping up my Ph.D. in electrical and computer engineering at Duke University when the Chief Information Officer approached my lab group to create a solution to ensure that the 1,100 cameras on campus functioned as they should.

It seems like a simple solution would be to just manually check every camera every day. But, for an institution with even hundreds of cameras, this task can be both tedious and subject to human error. So, we developed an artificial intelligence-based software that monitors security cameras and makes sure that they are all working according to a base/desired image.

I spun off this technology developed for Duke into a company, now known as Ai-RGUS. Since then, we’ve modified and added many components to our offering. Ai-RGUS is now a leading software solution which verifies that security cameras are operating optimally and alerts users automatically if a camera has a problem.

How has surveillance developed over the past decade?

In the past decade, surveillance has benefited tremendously from several technological advances. This includes the development of deep learning which has drastically improved the accuracy of artificial intelligence (AI) on imagery (among other modalities), the cheaper cost of computational hardware and the availability of higher internet bandwidths across the US.

These advancements have made new business opportunities economically possible. One example is the remote guarding paradigm to crime deterrence, where remote operators leverage AI-based automated event detection to monitor a much larger collection of cameras than was previously possible when human operators were continuously watching a handful of cameras.

These advances have also enabled camera system users to find specific information in their massive collection of surveillance videos. Rather than having people manually sift through the video, which can take hours or even days to accomplish, running the AI computer program conducts the task much faster.

How important is it to provide low cost, easy to use AI solutions?

Currently, one of the most effective ways to leverage AI is by using it to identify objects or events of interest. Highlighting if there is crowding within a certain area, detecting excessive motion somewhere you do not expect it or finding if a camera was tilted away from its intended view are all examples of AI doing its job.

Considering that a camera system’s fleet can grow to be hundreds, thousands or even tens of thousands of cameras, it can be very costly to have people continuously monitoring for irregular events. Using AI solutions divides the labour by letting humans inspect those events of interest and determine whether further action is required.

When such AI solutions are accessible, safety is improved because human operators without AI cannot provide the level of monitoring that is often required to keep people on their premises safe.

Maintenance is a critical, often overlooked part of the surveillance sector – how does Ai-RGUS make inspection and upkeep easy?

Ai-RGUS’ user interface saves operators’ time by streamlining the camera inspection process. Our software automatically alerts users about any problems with their security camera evidence they expect to have. We use AI to verify that security cameras are capturing clear images and are producing usable video evidence.

This includes automatically catching camera view problems such as: Problems due to blur, block, tilt, glare or low-light; camera/NVR/DVR misconfigurations or failures; wrong timestamp; missing or not enough days of recordings.

Users also get reports, alerts, statistics and even work orders to make correcting your cameras’ wellness easy. These key features include frequent inspection, providing a shortlist of unhealthy cameras for manual review, customised reports of camera and/or image issues and collecting historical camera system health statistics.

With Ai-RGUS, users are also able to implement a comprehensive cybersecurity program. With the software, security managers can remotely remediate an array of issues, including the ability to change out-of-date or insecure passwords, reboot devices and cameras as well as update firmware versions.

Daniël Reichman

Daniël Reichman, Ph.D. is the CEO and Chief Scientist of Ai-RGUS, an AI start-up spun out of Duke University. After receiving his Ph.D. at age 25, Dr. Reichman founded Ai-RGUS.

With over 24 university publications, Dr. Reichman obtained his doctorate in Electrical and Computer Engineering from Duke University from a program fully funded by the US Army Research Office. He also successfully completed the first two actuarial exams and obtained a minor in Mathematics.

This article was originally published in the March 2023 edition of International Security Journal. To read your FREE digital edition, click here.