So you will get frequent visits by the cops if you happen to live close enough to one of these systems! Or maybe you're just out walking the dog and happen to pass by 1 of these systems? Has to be some sort of illegal search thing going on there!
I want to know more. The video shows it picking up a gun that was pulled out and being used. What happens if I walk into a store that uses it while I'm holding a cell phone in my hand and point at something on the shelf behind the clerk saying can I have that blue lighter?
From the linked article:
"Athena offers multiple systems at a range of prices, but the one that includes gun detection, lock and elevator integration and real-time access costs $100 per camera per month."
Let's assume a school has welded shut all except three doors (ignore the fire hazard for a moment):
$100 x 8 months (conservative estimate) = $800 per school year, PER CAMERA. Multiply it by three cameras and you have to spend $2400 for that single school, per year, per camera. How many cameras would a real school need to cover the usual entrances? How many schools would a district choose to outfit with the cameras?
We keep hearing about how underfunded schools are and yet the most affordable SOLUTION - not a knee-jerk reaction, not a fancy band-aid approach - the most affordable and real solution is anathema to "them": let (volunteer) teachers be armed (and their identities NOT divulged to students) if they so choose.
Firearm: $300 - $900
Holster: $20 - $200
Training: $200 - $1200
Ammo: varies by caliber and frequency of training
Even taking their training costs into account (if paid for by the districts), a one-time expense with minimal upkeep (periodic training or recertification) is MUCH cheaper and places first responders on scene rather than wait for a fancy camera to call up the local PD.
The thing about AI is that even if it's not 100% at the moment it learns as it goes by being told what is and is not a legitimate positive. I find it completely plausible that by studying not just the shape of the object but also the human movements associated with it's use that this could be detected.
But still it's not likely to see broad use. It's too easy to mess with. A couple of kids having an airsoft dual is going to bring out SWAT. You could fool it for the purposes of creating a diversion. It's inability to distinguish between real and replicas is probably 0%.
And it died after the initial euphoria met reality. I predict the same thing happening here. As RustyYoungMan points out above, while $100.00 per camera sounds inexpensive when you factor in the number of doors per school to be guarded it becomes a significant budget item.
And its implementation is naive, to be polite. A computer-generated voice asking the bad guy to put down his gun because the police have been called. Really? According to the accounts I have read lately, the majority of people doing these shootings end up dead, either by their own hand when the police arrive or when the police shoot them. But it's after they have done their bad thing. So at best this technology will alert the police a few seconds before the first shots are fired. Hopefully if it's considered at all, reasonable people will see that it does nothing to prevent shootings and will save their money.
Fishinkeylargo also brings up an interesting point, a false positive. Not far-fetched given a poorly lit scene or one where the camera sensor is overdriven by too much light. And there's the subject of false-positives. Clicking the "artificial intelligence" link in the OP story takes you to this:
What's really interesting to me is all the technology coming out to help "prevent" shootings. None of these devices can prevent anything. Like she said most cameras just record events. Well, isn't that what theirs does also? The only difference is it sends an alert somewhere. Saw a company this morning that developed a table that can be turned to create a barrier to hide behind (bullet proof). All these devices are good in conjunction with some simple tools. Oh, I don't know, maybe locked doors, situational awareness and firearms on premisses? It's great that the gun is detected or you have something to hide behind but what next till the police show up?
This is the kind of thing that could lead to some kind of jump ahead in holster technology, the building of a 'stealthy' holster that has different angles and such where it couldn't easily be detected under ideal circumstances. Who knows.
People are always trying to invent novel ways to conceal guns. I simply cannot subscribe to their innovations. The natural point from which to draw is on your side so that raising the arm the hand can reach the gun easily. There is reason why belted holsters since the invention of the pistol holster have been at 3’o’clock. Your arms is at your side and you raise your hand to find the holster and gun in one swift motion.
Of course that that does not address concealability. If carryin IWB, which I do, I carry at 4 o’clock. For me that makes the gun adequately concealable and immediately accessible. I do not wrestle in my mind with ankle, appendix, etc carry. I want to be able to present my gun as quickly as possible without any excess body movement I know that in a real gunfight I’d I have to reach over my shoulder to access my pistols will likely be dead or wounded before I can even aim at my assailant.
When concealment becomes the primary objective to the concealed carrier best practices for self defense takes second place. That is self defeating. I do not open carry unless hiking or camping. Otherwise I carry concealed. I have zero interest in advertising that I have a gun on my person. However, if my choice is to place the gun readily accessible or less accessible I will adopt the readily. If someone can tell I am carrying so be it. I concealed carry legally, and I do not need to try to hide it.
There was nothing in the video or the write-up to indicate it can spot a concealed weapon of any type. It can only see it when it is out in the open and they are including activities such as fights etc. Therefore, we do not need to worry about this technology at this time unless you pull your gun out into the open.
I don't see this as 99% effective as they claim, at least as far as CC goes. As Tankercorps71 pointed out, it might be easy to jump to the conclusion they're talking about concealed carry, but that's not specifically mentioned. (I think the lead photo in the article was used to deliberately foster that impression.) AI has gotten to the point where I can believe it can be trained to spot a gun in the open but if the technology is based on visual cues only, there's no way AI can spot a concealed weapon any better than a human.