In the FBI’s Crime in the United States, 2015 report, the intelligence and security organization revealed some startling facts about robberies in the U.S. According to the data provided, there were 327,374 robberies that took place in 2015 accounting for losses of $390 million (at least for those robberies that were reported).
While that number feels shocking enough, it’s the further elaboration on the weapons used during those robberies that provides a lot more insight into how these crimes are pulled off:
- 127,053 used strong-arm tactics
- 123,358 used a firearm
- 23,759 used a knife or other sharp object
- 27,397 used a weapon of some other type
With 46% of robbers relying on weapons to pull off their crimes, one way to quickly catch a threatening event is to catch the presence of an instrument commonly associated with aggression. Other key physical indicators, like the presence of a mask, or simply the presence of an unrecognized, unexpected visitor, can also be captured to quickly identify suspicious or immediately risky individuals on a business’s premises.
New Jersey-based, Deep Science AI has developed a technology that merges artificial intelligence with current business surveillance and security systems in an attempt to spot and prevent crime. While technology like AIsight–installed around Boston after the Marathon bombings–attempts to identify potential criminals through recognition of things like, “Casing the joint, poking around where he shouldn’t be, going around looking at the back entrances to buildings”, enhanced object identification technology is at the heart of Deep Science AI. Their goal is to offer a more comprehensive and efficient real-time security solution to businesses that might not otherwise be able to afford it or manage these threats on standard budgets with standard security staff.
According to their website: “We leverage the incredible progress recently made in computer vision thanks to a type of AI known as deep learning…Our AI Surveillance (AIS) platform can spot weapons, people concealing their identity, fire, and intruders after-hours, all at a fraction of the cost of a human-only monitoring solution.”
Using Deep Learning to Fight Crime
Up until now, Deep Science AI was only offering their security monitoring service in the United States to a small set of customers. There were originally 18 locations enrolled in the Beta program to test the technology in their stores. That number has recently jumped up to about 200 locations. In the future, Deep Science AI intends to open their service up to Canadian customers as well.
This new security monitoring service has proven to be beneficial for small businesses that can’t afford security teams and expensive monitoring services. Surprisingly, larger businesses have also shown an interest in Deep Science AI as it is an affordable supplemental service that can be added onto their existing infrastructure.
Pricing currently ranges from $44.99 to $59.99 per camera per month, and includes 24/7 monitoring, storage of the video feed containing the recorded robbery, free AI platform updates, as well as priority support provided by Deep Science AI.
In addition to providing businesses with an affordable solution to monitor their company’s premises for real-time threats, Deep Science AI also has a team of analysts who monitor their customers’ security camera feeds 24 hours a day, 7 days a week.
“The accuracy of deep learning techniques allows a single analyst to monitor hundreds of camera feeds, rather than a dozen…. Any detected threatening event is handled by one of our monitoring analysts, who immediately contacts your local authorities when the situation calls for it.”
For businesses that worry about false alarms (since no AI system is going to be right 100% of the time, at least for now), Deep Science AI enables customers to define the follow-up protocol by which analysts must handle their business’s detected threats. So, analysts can either call the business owner to inform them of the suspected robbery or they can contact the authorities right away.
By keeping their security solution affordable and flexible (there’s no hardware purchase required; just the monthly camera monitoring fee), Deep Science AI is showing that a big field like deep learning has the potential to bring about big changes to the small business landscape.
Brenda Barron is a writer from Southern California specializing in technology and business. When she’s not writing, you can usually find her outdoors with her family or knitting.
hello
If this software can really detect what is being claimed, that is a major advancement in the technology and home security world, isn’t it?
This is brilliant, on paper. At this point, I have heard about a few too many incidents involving AI and how it can be a very bad thing for people.
So this form of tech can be added to an existing security system? I think that is a major selling point!
James, yes, no additional hardware is needed. Just enough bandwidth to remotely watch the store’s IP cameras.
Oh, I can see this really is a benefit for those against crime. I think that having an AI watch for things like guns or a knife or even escalated behavior is a great start to keeping crime low.