What is Digg?66 Comments
- jbeldock, on 11/11/2009, -0/+18Hi, gang. James from ShotSpotter here. Very interesting comments; thought I would add some clarifications and facts to the list. (On a personal note: why is it that every DIGG entry ends up with a million negative comments saying "it's junk; I hate it; booooo!" In our case, ShotSpotter is deployed in 50 cities nationwide--not very likely that it would be "useless" with that kind of track record :-)
1. Big Brother stuff: definitely a question often asked. For the record, the system DOES NOT trigger on human speech or any other non-"impulsive" sound. "Impulsive" means loud and "bang like." Mathematically, that means the waveform displays a very rapid onset and a pretty typical exponential decay envelope. So we're not going to be listening in on speech. In terms of whether it's "big brother" to detect gunfire at all, bear in mind that the act of firing a weapon is illegal in the locations in which we're deployed, and the sound is made publicly (i.e., one has no reasonable expectation of privacy when firing a weapon which can be heard for roughly a mile in each direction outdoors).
2. Aquarius has it right that the vast majority of gunfire doesn't get reported to 9-1-1. Depending on the city we're deployed in, the reporting rate is anywhere from 11% to 70%. All but one of the 50 cities which have ShotSpotter systems have reporting rates below 50% (in other words, only one city has that 70% number). On average, when a 9-1-1 call comes in, it comes in 2-3 minutes after we detect it and notify dispatch. (San Francisco, for example, has a 16.7% report rate and an average of 2:35 minutes delay on those reports. Delay is due to all sorts of things: finding phone, deciding whether to call 9-1-1, cell phones being routed to state police or CHP in California, etc.)
3. The pricing data isn't right. That's odd, because our pricing is a matter of public record in most cities. The system costs less than half of what people are saying (i.e., it DOES NOT cost $500k/square mile :-). Also, to be perfectly fair, to the people asking about CCTV: generally ShotSpotter is deployed ALONG SIDE CCTV. It helps solve two critical CCTV problems: (a) if the camera is pointed the wrong way, then obviously CCTV doesn't capture an incident; (b) with 100s if not 1,000s of camera feeds coming into a given command center, there's simply too much data. ShotSpotter is usually used to (a) point the cameras automatically; and (b) highlight which of the various cameras should be watched. In general, cities spend anywhere from 3-5 TIMES as much on CCTV as they do on ShotSpotter.
4. Regarding false alarms: yes, there are false alarms. No more than with 9-1-1, to be certain, but bear in mind that we're detecting sometimes four to five TIMES as many incidents as get reported to 9-1-1, and that naturally increases the numbers of true and false positives. But to the person saying "9 out of 10 incidents is a car backfire," that's just not accurate. We have the statistics for that system, and the false alarm rate there is roughly even with the rest of the country. (Also, in terms of false alarm: compare the impact of one of our false alarms--the dispatcher clicks on our incident, listens to it, and then clicks "acknowledge" and it goes away--with the work required to track down a 9-1-1 false alarm--dispatch a patrol vehicle, drive to the location, spend 10 minutes looking around, etc.)
5. Somebody said our salespeople are suggesting that police circumvent dispatch and go straight to patrol or local commanders. That's also just not accurate. (Interesting: someone said NYPD and NY State Police think this is true. Neither of them is a customer, so I wonder how they got this "fact"??) Anyway, our primary user interface software runs *IN THE DISPATCH CENTER* and is used by Dispatchers. We also have a software package for use in patrol vehicles, but in every customer account I'm aware of, there's a very careful coordination between dispatch and on-scene commanders, and usually the on-scene dispatch is only active for limited, tactical purposes.
6. Finally, I can offer a few more facts: First of all, we've had numerous customers point to 30-40% reductions in both violent crime and homicides after deploying our technology. Is that *because* of ShotSpotter? No, it's because they (the police department) have better information and use us as a tool to do their jobs more efficiently. Second, since we started counting, we've helped police and first responders locate 220 shooting victims nationwide and render life-saving aid. Third, there are a number of cases of officer-involved shootings where our evidence has been integral to clearing the officers vis-a-vis appropriate use of force. In the end, we're just a tool, like any other technology used by police, but calling ShotSpotter "useless" is probably a little unfair. :-)
-James - EddiePotato, on 11/11/2009, -2/+19I was curious how long it would take for a Big Brother comment. Not long.
These things are awesome, and should be in every city prone to violence. There is no valid or legal reason for anyone to be shooting a gun outdoors in an inhabited area, unless it's self defense, and even in that case cops should be notified immediately. - Aquileria, on 11/11/2009, -3/+15Whoa, weird, we were talking about this in my criminology seminar yesterday; this technology is useless and state police & NYPD officers in our class were openly mocking it. They have this in Troy, NY and Rochester, NY and all it ends up doing is become a glorified tally for how many gunshots occur in a given area... It costs 500,000 dollars per square mile of coverage (which could pay for dozens of standard CCTV cameras + employees to monitor them and alert dispatch).
The amitious plan that salesmen pitch to police chiefs is that these cameras could cut out dispatchers as middle men and make it so police can directly respond to incidents, but it doesn't happen and it wouldn't matter anyway. Research shows that whether police response times are 3 minutes or 3 hours, they won't be able to interdict the crime at all.
Basically, sorry to say because it sounds cool, but this is a huge waste of municipality money... You'd be way, way better off doing what NYPD does, which is simply going into stores + banks that already have private security cameras installed, and analyzing their tapes during investigations. Let businesses pay for the cameras instead of wasting already limited police funding. - strictnein, on 11/11/2009, -0/+7Minneapolis gunfire maps:
http://www.ci.minneapolis.mn.us/police/crime-stati ...
They're PDFs created weekly. - EddiePotato, on 11/11/2009, -0/+7These maps are also a good way to find affordable housing.
- ricksite, on 11/11/2009, -0/+6How long until they update the system to shoot back at the source of the gunshots?
- captininsanity, on 11/11/2009, -1/+6At first though a microphone would seem less invasive than a camera. Also you shouldn't talk about sensitive content in public anyway. However, as with all surveillance systems this has the potential for abuse. On its own its not that bad, but for some its a slippery slope, especially if we start talking about adding in voice recognition and speech recognition software to scan speech for suspicious conversations. The kind of thing that people are paranoid about with wire tapping. Legally though anything you say in public is fair game to record for legal purposes...
- LostSoul83, on 11/11/2009, -0/+5Didn't they have tech like this in Deus Ex? I seem to recall Jaime telling the player about this in the training mission.
- AndrewDB, on 11/11/2009, -1/+4I love the physics of sound waves.
The case of longitudinal harmonic sound waves, the frequency and wavelength can be described with the formula: y(x,t) = yo sin (ω(t-x over c)).
where:
* y is the displacement of the point on the traveling sound wave;
* x is the distance the point has traveled from the wave's source;
* t is the time elapsed;
* y0 is the amplitude of the oscillations,
* c is the speed of the wave; and
* ω is the angular frequency of the wave.
The ordinary frequency f, in hertz, of the wave can be found using:
f = ω over 2π..
----
So, wouldn't it be possible to see where the signal bounces off of buildings where the sound is "heard", and triangulate where the shooter was at the same of the time of the shooting? - drethedog, on 11/11/2009, -0/+3The cops could use the SWATS system in the patrol car.
EARS/SWATS is a small, compass-like device that a soldier can wear on his arm or that can be mounted on a vehicle. When a shot is fired it provides almost instant information either on a visual display or as an aural feed on the precise source location of that shot. - Gedd75, on 11/11/2009, -0/+3Yes, in the Hong Kong mission. What made them effective there was the on-guard, heavily armed, Mil-Bot's just waiting to fire back at anyone firing a weapon. So, yeah, gotta agree with one of the posters above, these AGS's are useless on their own. At least, until ED-209 is on patrol with them.
- srodolff, on 11/11/2009, -0/+3http://www.thefreelibrary.com/Truck+crews+test+ant ...
- SuperToaster, on 11/11/2009, -0/+3I wonder if officer's see little arrows pointing towards the source of the gunfire in their HUD ala Halo
- Leopards, on 11/11/2009, -0/+2If it works as shown in the video this would be great for a lot of cities that have a gun violence problem! But remember this is put out by the Seller, so a grain or two of salt is recommended!
- srodolff, on 11/11/2009, -0/+2The military already has that........
- Aquileria, on 11/11/2009, -0/+2The stars are the shooting, the cameras nearby activate and take a photo, even if they're only staring at sidewalk or a side of a building not facing the incident or even on the same street as the incident. Now this isn't totally useless, crime analysts can use this kind of thing and perhaps on rare occasion spot a common denominator in shootings like a vehicle, but keep in mind these cameras will go off even if what they "hear" is just a car backfiring! That's likely the reason there are activations of ShotSpotter far away from the reported/investigated shooting incidents.
Also, there could have been actual shots fired at those far-off activations, but the majority of shooting and crimes in general are not reported to police (usually because victim/target + offender are BOTH up to no good [e.g. drug disputes]) so they can't investigate what isn't reported, or plot it on the crime map. - feezus, on 11/11/2009, -0/+2It's been a while since I played it, but didn't all of the thugs in Hong Kong carry swords because of this?
- Antialias, on 11/11/2009, -0/+2I remember hearing about something like this the military uses to spot snipers. With echos and a tense situation it's hard to know where a shot came from.
- shapedy, on 11/11/2009, -1/+3See Digg? Not all cops are bad.
- ImTheDuke, on 11/11/2009, -0/+2They've had this in Oakland for a while
- HonoredMule, on 11/11/2009, -0/+2"...maybe you're just from a place with a lot of selfish, *****, self-absorbed, corrupt people?"
We just call it Earth for short. - AndrewDB, on 11/11/2009, -0/+2I love it when everything comes back to mathematics.
- EddiePotato, on 11/11/2009, -0/+2Yeah, they are in the Hong Kong missions. Fire one shot and those giant security bots and military police will tear you up. But I believe the idea was based on the fact that they already have these things in Hong Kong (and did back when Deus Ex was made).
- rmxz, on 11/11/2009, -0/+2Hi James, Some replies to what you wrote:
"On a personal note: why is it that every DIGG entry ends up with a million negative comments"
As you notice - every digg entry, not just shotspotter ones - attract a lot of skeptics. Lots of people/companies tend to spam Digg with advertisements disguised as articles. As a reaction, it seems people post objections, and other people reply with defenses, and the articles which are well defended in the comments tend to rise up faster.
"not very likely that it would be "useless" with that kind of track record ":-)
Though you guys do have some excellent salespeople too. :-)
"Big Brother stuff"
Seems digg has a pretty big contingent of people who interpret the second amendment as attempting to empower people relative to the government; and technologies that limit that empowerment are often criticized on digg.
"Second, since we started counting, we've helped police and first responders locate 220 shooting victims nationwide and render life-saving aid. "
Very nice! Perhaps your PR guys could highlight this statistic more to get more support from the general public - especially critical audiences like online communities like this one.
"Third, there are a number of cases of officer-involved shootings where our evidence has been integral to clearing the officers vis-a-vis appropriate use of force."
Does this work the other way too - any cases where shotspotter evidence contributed to implicating someone for inappropriate use of force? Might play better with the audience on Digg. - diggbigwig, on 11/11/2009, -0/+2I know the military uses this method but hasn't this been in place in other cities (LA?!?!) for a long time.
- LostSoul83, on 11/12/2009, -0/+1Yep, I think only the police/security guards had guns in that portion of the game. The guy in front of Tong's compound had a combat shotgun though. One time I got him into a battle with the above mentioned security bot. He is invincible, so he (eventually) won the fight.
- sasquach111, on 11/11/2009, -1/+2I have a friend in the Troy police department and she tells me that 9 times out of 10 the thing registers it's a car backfiring. Apparently, it takes a year before they can calibrate the thing to tell the difference. It's a pretty big waste of taxpayer money IMO.
- AgentVladimir, on 11/11/2009, -0/+1Huh I never put the two together but I think you might be right.
- strictnein, on 11/11/2009, -0/+1I believe the red circles with no dot are 911 calls. The red circles with dots are the gun shots detected by the ShotSpotter system.
Per the note at the bottom:
"One incident may be represented by multiple Icons in the map, depending upon the number of callers and location of incoming calls."
The stars:
"Officers are dispatched to the location where the shooting victim is currently at, which may include the hopsital, or location different from the actual shooting incident."
Looking at some of the maps, many of the "shooting" locations are simply the major hospitals in the area. - jbeldock, on 11/12/2009, -0/+1Brown2hm: James from ShotSpotter here. We do all sorts of interesting non-triangulation localization math as well as triangulation and angle-of-arrival math. Would be interested to talk to you about your work. There may be some potential cross-applications or perhaps you've got something we should license. Send me an email to jbeldock (at) shotspotter (dot) com.
-James - vicsvenge, on 11/11/2009, -0/+1They just added this system to east side Saginaw Michigan. I think it's a good technology, hopefully it pays off. I doubt anyone who is going to shoot someone really cares if someone is listening or not. Makes a pretty big bang...
- shipwreck58, on 11/11/2009, -0/+1This is not news. The Boomeraing counterfire system has been used by the Army for a few years now to try and detect snipers.
- bluesman3535, on 11/12/2009, -0/+1Gov-speak terminology is pretty damn confusing. 'Acoustic sensor'? Oh a spy microphone! Got it.
- EddiePotato, on 11/11/2009, -0/+1They should fire interceptor bullets, to take down the bad bullets mid flight.
- mrmaple, on 11/11/2009, -0/+1Cool, but what's the math behind the ability to characterize each sound? It sounds like AI, but WWII submarines could also characterize the sounds around the submarine (like other subs.)
- and303, on 11/11/2009, -0/+1Um, in a major city they are.
- kmvtech, on 11/12/2009, -0/+1Yeah, this read was so curious like any other suspence novel.. well mixed!
Thank you for nice post! - CamonDraconis, on 11/11/2009, -0/+1Your own firearm can still help defend you in a matter of seconds whereas cops still take upwards of 5 minutes.
- CamonDraconis, on 11/11/2009, -0/+1Wow... look at how inaccurate that is.
- brown2hm, on 11/11/2009, -0/+1It is possible. I'm actually doing my Ph.D thesis on non-line-of-sight source localization techniques. Triangulating via reflections of sound waves seems to be a good idea, but it's actually pretty inaccurate in practice.
The techniques we use instead are pretty impressive in accuracy, but also computationally intensive. I've also have ways to reconstruct the original sound at the source while only hearing echo's from behind behind buildings.
Math is cool. - AndrewDB, on 11/11/2009, -0/+1That, I unfortunately, don't know.
I'd assume that each weapon/gun has a distinctive sound in frequency when it "rings out", and given that, you can count the type and time of oscillation or oscillations. - and303, on 11/11/2009, -0/+1Not newsworthy. This has been used widespread in Chicago since 2004.
- Aquileria, on 11/11/2009, -0/+1Funny you should mention Robocop... That movie, to me, always seemed to put out the best argument for how to improve policing: just augment our already good police officers with the best technology to keep them alive, and give them public support (like they do in Japan). Do this as opposed to putting out ridiculous police-replacement robots that end up going on killing rampages and instead of privatizing police (the mercenaries in Robo 3)
I mean, I dunno... People on digg complain a lot about the police, but I don't really understand it, 99% of the law enforcement people you guys interact with are FROM your community, we don't have overbearing national police, and state government police forces are a lot more hands-off and really primarily just patrol the highways. And even they are divided into troops, so they're from your local area and community too! In other words, maybe police aren't the problem, maybe you're just from a place with a lot of selfish, *****, self-absorbed, corrupt people? lol. - and303, on 11/11/2009, -0/+1Thank you for your comments James.
It is an extreme rarity that anyone posts a comment and knows what the heck they're talking about. I think your system is an innovative counterstrike in neighborhoods plagued by violent crime.
Most of the people bitching about it do not live in an area that benefits from it. - feezus, on 11/11/2009, -0/+1"So, wouldn't it be possible to see where the signal bounces off of buildings where the sound is "heard", and triangulate where the shooter was at the same of the time of the shooting?"
I can tell you that without all of the math. The shooter was the guy mugging the vicitm. It's not like people are camped up in high rises sniping people. - martoq, on 11/11/2009, -0/+1This was developed in England and them marketed to municipals in CA. A city near Albany, NY was one of the first cities in the country to test it and it has been extremely successful in what it is trying to accomplish.
- Aquileria, on 11/11/2009, -0/+0CCTV's/monitoring are all about evidence collection and crime deterrence, you can't neutralize criminal acts... The offenders are long gone by the time police arrive most of the time, except in movies.
- m00nmaster, on 11/11/2009, -1/+1Newark, NJ does this and I'm sure other high profile cities do. If not, they should as they are extremely accurate and the shooting/murder rate has gone down significantly since they were installed in the more violent prone areas.
- pleased2Bhere, on 11/12/2009, -0/+0If it's been around for so long this would explain why this technology hasn't become more widespread.
-
Show 51 - 67 of 67 discussions
© Digg Inc. 2009
— Content created and posted by Digg users is