Police Gunshot Location Detection Systems
|
A post by "Stacks " To see more posts click here |
Last Sunday’s Washington Post featured a little four line blurb in the Metro section about a ‘Body Found With Gunshot Wound’. It occurred late on a Friday night in northeast Washington DC and while I feel bad for the family that lost a loved one, this incident is not anything shockingly new for the area. What caught my attention was how the police became aware of the shooting. According to the blurb:
“Authorities went there after an alert from the police gunshot sensing system.”
Now I may very well be alone on this, but up until now I had no idea such a thing existed, let alone how it worked, how much it costs, how many of them are out there, and finally, why is this the first I’m hearing of it? As it was just a blurb and no further information was given; I decided to research it a little further.
A quick Google Patents search turned up a description as to the basic idea behind it. From what I am able to discern, these are the essential principles behind the system:
- A number of communication poles are aligned throughout a grid (picture an intersection or possibly multiple intersections in DC)
- Each of these poles contains microphones (to of course pick up sounds), a data acquisition and signal processing unit (to filter out normal street noise and also tuned to detect both the range of frequency & the duration of the pulse that is created by a gunshot), along with a power supply, and an external communicator (to relay the information to the authorities)
All of that sounds pretty cool, but technologically speaking, it doesn’t seem to be much more complicated than your standard glass-break detectors that are staples of your standard home security systems.
According to the search, the patent was filed in the middle of 1996 and issued at the end of 1999, meaning that this technology seems to have been around for about 10 years now so…is their use widespread?
Another quick search turns up a news article from less than a month ago in Richmond (CA), wherein police credit gunshot detectors for pinpointing the locations of 3 separate shootings in just one week.
The system on Tuesday successfully led to police to a home where officers found a sawed off AK-47, a semiautomatic pistol and a handgun and then the following day police responded to 2 more incidents:
“A spokesman for the company [that makes the sensors] noted gunfire sometimes goes unreported in some communities where shootings are more common.”
Okay, that’s troubling…I wonder how frequently shootings have to occur for that to happen…
The aforementioned article states that the city spent $600,000 to install 16 sensors per square mile (the city of Richmond, CA has a landmass of 30.4 sq miles) throughout the city. It also states that if a gunshot is detected by 3 of the sensors, that it can triangulate a location accurate to within 25 meters (roughly 82 ft or 27 yards for those that need a football reference to visualize it). I am sure that not every square inch of landmass in the city is covered by these detectors, if we estimate 25 sq miles (and 16 sensors per square mile), we are looking at a very rough estimate of about $1500 per sensor ($600K for approx. 400 sensors).
In DC, while I can not verify the number or the locations, according to this Washington Post article from 2006, they are (or were at the time) in use, complementing the previously existing surveillance cameras.
On the surface this seems like a lot to you or me (and my estimates could very well be completely inaccurate), but if you look deeper into it, it does not seem to be an exorbitant amount to pay when you consider the potential benefits that it can offer. My initial thoughts assume that it would create a quicker emergency response to the victim (and thus save lives) and would also increase the chances of catching the shooter (both in the improved response time and also in the ability of it potentially corroborating eyewitness testimonies). Additionally when you factor in how much government normally overspends on pointless projects, this seems to look like more and more of a bargain.
I do have a few concerns with this technology. The first is what was referenced earlier. If gunfire is going unreported “in areas where shootings are more common” it means not only that it won’t be effective where it is most needed, but it would also create even more of a disparity between the rich and the poor. They work in Georgetown, but not in Southeast? Wouldn’t that even encourage criminals to ensure that the high-crime areas stay the high-crime areas so that they have a higher chance of getting away undetected?
Secondly, I have a documented distrust with the idea of just putting everything in the hands of the government and assuming that they’ll only do good with it. If we have a city with 16 sensors per square mile and every sensor is equipped with a microphone as well as the ability to differentiate between different frequencies and sounds, who’s to stop these device from potentially being equipped to become a network of listening devices.? I am by no means a conspiracy theorist, but this raises some very serious privacy concerns.
Okay so maybe I was a little off with my initial comparison to your standard, antiquated glass-break sensors. This equipment has the potential to do many positive things, but at what cost? Since I have yet to have time to really solidify my opinion on this, I am very interested in hearing your thoughts on this. Are there any major issues that I’m not considering?
Here’s a final little bit of information to help with your judgment. I found a brand name for a company that sells these sensors. They have a FAQs page, but keep in mind that this is a company that is trying to sell their product, so they aren’t going to be without bias. Here is their statement about the Big Brother question that I raised earlier. While it does make the (valid) point about the fact that “it is illegal to discharge a firearm except on a certified shooting range or under other controlled circumstances, thus, anyone who does fire a gun has broken the law”, it completely (and intentionally?) ignores the points I raised earlier about privacy issues for those not breaking the law.
Form an opinion, let me know your thoughts, and
udothedishes….
[To read about how the military uses this technology to protect our soliders, check here and here.]
One step closer to total surveillance. Next, these same noise detectors will have a camera that comes on when a shot is detected. Everyone out there should watch or read 1984 by George Orwell.
I think the point of these devices is well intended, but as usual and as you point out, the potential for abuse to our privacy rights is very high with this. Add to it the fact that these devices were implemented without much public notice tells me that someone didn’t want too many people to know about them.
Maybe we should give the people over at the ACLU a call and see what they think.
Hi, gang. James from ShotSpotter here. That’s our system in Washington, DC as well as 45 other cities nationwide (including the system in Richmond, CA). Let me add some data and maybe address a couple of your concerns:
First of all, the firmware in our sensors *DOES NOT* trigger on anything other than gunfire (or other loud, impulsive noises, such as explosions). They *DO NOT* trigger on voices, and the government has no ability to adjust the firmware or otherwise change what they trigger on to make them “listen” to voices. They also transmit only 4 or so seconds of audio when they detect gunfire, which obviously is not enough time to eavesdrop meaningfully. Contractually–this is all a matter if not of public record then available by FOIA request–the government owns the gunshot data and audio from the incidents, as well as the sensors, but they do *NOT* have the right to manipulate the system directly, to change its audio recording behavior, firmware, etc.
Second, the photo and patent you found is of a competitor’s system, not ours. Their system is deployed down at light-pole level, and in fairness I believe it doesn’t trigger on anything but gunfire either (i.e., not on human speech), although I don’t know for certain. I can see why the placement of their sensors (down at light-pole level) might have given rise to this concern. By contrast, our sensors are deployed on the tops of flat-roofed commercial buildings, usually the highest in the area, and therefore are way away from the street. (We get better performance by “listening to the horizon” up at that level.) The microphones inside are similar to (actually, the same physical electronic parts) as cell phone microphones. The result is that speech is only intelligible for a 20-30 foot radius (maybe 50 feet, to be generous) around the sensor. Which is on top of a roof, maybe 300 feet from the street. Which only transmits audio for about 4 seconds. And only when there’s a gunshot.
Now, about the demographic/crime reporting questions you raised: the numbers aren’t mine, they’re from our customers, but what I can report is: San Francisco has a 17% reporting rate (this is public record). In other words, they get a phone call to 9-1-1 approximately 1 in 6 times that we detect gunfire. When they do get the call, it comes in on average 2 minutes and 35 seconds after we trigger (which is roughly 5 seconds on average after the physical trigger of the gun is pulled). You have guessed correctly that the higher crime neighborhoods have lower reporting rates, but that is absolutely *not* unique to gunshot detection: it is a fact of crime sociology life. That doesn’t eliminate the concerns you’ve raised, but it does put them in some context.
About costs: this is realy money, and obviously I have a business to run. Having said that, let’s put some numbers around *not* dealing with gunfire. University of Chicago study shows that the lifetime cost of a shooting to society (medical, investigatory, incarceration, etc.) is $170k (that’s a *SHOOTING*, not a homicide). Cities deploying our technology see anywhere from a 30-40% reduction in violent crime and sometimes as high as a 70% reduction in gunfire rates. So you can see almost immediately how there’s cost savings to the city. (Yes, that’s a real Return On Investment, even if we are tracking broadly Societal Cost, which obviously doesn’t come out of one specific departmental budget.) Second, depending on the city, a homicide can cost upwards of $1m to investigate. Los Angeles has seen a 40% reduction in homicides in our coverage area (again, their numbers, not mine). Further savings.
Finally, on lives saved: it does happen, but maybe not as often as we’d like. Bear in mind that the usual 9-1-1 caller cannot usually provide a precise location. We are accurate enough to say “behind the house, near the bushes” (we are using Google Earth-like imagery–no invasion of privacy there
. So we have gotten police to the scene of someone who might otherwise have blead to death. Wish it happened more often, but then again, if it did, there would be more people shot, which I don’t wish.
So, hope that helps. Happy to keep replying if others have thoughts/questions/comments.
Best,
-James from ShotSpotter
I completely agree that while it may help in crime solving/prevention, it could potentially be a threat to privacy. There’ s no telling that with this technology being implemented on our street corners is being set up with the facade of “crime solving/prevention” when it’s really meant to be spying on the public. “I have so many doubts.” George Orwell was psychic, it’s all coming true.
Good article but misleading in the Georgetown comment, the shotspotter sytems are not placed in the most expensive nieghborhoods; although the new Shaw one does cover adams morgan. What I do find “Orwellian” is that a company spokesman found this article so quickly in a local blog.
James, please don’t take this as a knock to your company or product as i do agree that you guys are helping out a lot with your technology.
However, just because the gov’t doesn’t have the *right* to manipulate your system, doesn’t mean they can’t and won’t. Wiretapping is illegal, as legislatively mandated BY our government,…yet we all remember Watergate, don’t we?
In addition, I’m an electrical engineer by trade, and we learn our trade by understanding technology and then reverse engineering things, so I’m hardly convinced that they can’t and won’t, or more so don’t have a reason to, manipulate your device.
So although i’m sure you sell your device in good faith, it can be used for much more with a little tampering from someone with an engineering background. After all, the gov’t has enough money to pay people to do these type of things bc they tax the **** out of us.
And furthermore, just because you guys sell only sound-related devices (that capture 4 sec of sound), that doesn’t mean your competitors aren’t coming out with bigger and better things. In the city of Philadelphia they use similar type crime technology, but from what I’ve seen and heard theirs is camera and sound technology, AT lightpole level. Nice and close for whoever’s behind the camera to watch…
At the end of the day, it is simply a step towards social control, regardless of how you paint it. Even if the devices never get tampered with, they will be an argument that any lawyer can use to argue a case for the next more invasive thing to install and wire up.
With all that being said, my criticism is more of how your device will be used, rather than the fact that it was made. All good things get used first, then abused later. Just read our history books…
P.S. Just some FOOD FOR THOUGHT….I read a statistical estimate a while back that i’ll share…
The people born in 2049 or later will have their ENTIRE lives video recorded in one form or another. Be it willingly thru family videos at events such as birth, to security cameras in stores, to redlight cameras, to street-level crime cameras (City of Phila.) etc….
To put it in perspective,..that is when us twenty-somethings will be retiring….so basically our grandchildren.
James from ShotSpotter here again. Pavla, let me clarify my statement: I know for a fact that there are no government agencies using our devices to eavesdrop on conversations in DC, Richmond, etc. I know this because we are contracted to control the sensor network, and the state/local government *DOES NOT* have authority or access (password, encryption keys, or the knowledge needed to break into the system). We take security pretty seriously: we’ve got robust intrusion protection, including not least among which are encrypted firmware and write-only DSP code. We also ping each of our sensors, nationwide, on a minute-by-minute basis. Among the telemetry data we get back is firmware rev and MD5 hash of firmware image. And the round-trip comms are on application level encrypted channels in a proprietary format that only we know how to parse. The sensors therefore are running code I *KNOW* is ours, and I *KNOW* what they do: they only trigger on signals that have the fast rise-time and exponential decay of gunfire or explosions.
Is it possible that the NSA or some other agency could break into our sensors? Absolutely. They can do virtually anything they want to, I’m sure. But that’s not the same as saying, for example, that a local police department has those same resources. They simply don’t. Moreover, if NSA or CIA or some other agency worthy of conspiracy theories wanted to eavesdrop on what you and I are saying, legally or illegally, they most certainly don’t need ShotSpotter’s help (nor, for the record, have they ever asked for it
). They have far more sophisticated devices capable of hearing human voices at far greater ranges than the 20-50 feet on top of a rooftop at which our sensors might be able to hear intelligible voices. For all we know, they can force your cell phone or my cell phone to listen for them. That would be a far better way to hear meaningful conversations than to go after ShotSpotter sensors which are located in the perfect position to hear gunfire (on rooftops) but in a lousy position to hear people.
Just to clarify on one other point: they system in Philadelphia is a ShotSpotter system. It is not integrated with the Philly camera system.
Also, to the person who thinks it’s frightening that “we” found out about this blog so quickly, I guess I’m a little surprised that it is frightening. Google has a service called “Google Alerts” which tells anyone who wants to know about any keyword which pops up on the Net. I subscribe to it for the word “ShotSpotter,” which is how I found out about this blog. Is that really a sign of Big Brother, or simply a sign of a more connected/rapid-fire world?
Saving lives/money and Crime prevention vs. privacy? I for one believe this is a technology put to good use and wouldn’t mind seeing more of it being used across a wider range of cities!