Wasn't the Shot Spotter thing pretty useless in reality though?
Depends on what you define as "useless".
It was objectively bad at identifying gunshots (the precision and recall were horrendous) and at reducing gun violence.
But it was really good at manufacturing probable cause, because the false positive rate was so high that police could use the sound of a dumpster lid closing as a legal reason to detain and search anyone who they wanted in that general area.
https://www.eff.org/deeplinks/2021/07/its-time-police-stop-u...
I live near a lot of urban gunfire and I can't imagine what they would do unless there is a dead body.
What if the arrival of the police or the sound of sirens dispelled the gunfire early enough to save those bystanders?
What if that body they find isn't quite dead, but bleeding out and they are able to save them?
To the point where suggesting officers should respond incase someone is bleeding out get downvoted.
I guess the compassionate thing is to wait for the body on Monday because "what's the point"
1. Administer first aid
2. Gather witnesses
3. Preserve evidence
4. Coordinate with other units to grid-search for suspicious persons or vehicles
5. In some cases, gather evidence/testimony that enables immediate pursuit
You know… police work.
However, at some point there are too many shots to even show up for, let alone do the more involved tasks you mention, and I think this is the situations some cities using the service found themselves in. At that point, it just becomes a data collection project, with unclear goals.
For what it is worth, I think we have ShotSpotter in my location, but have never seen police physically responding to shots fired. That is not to say it isn't useful in other locations
Having the resources and political will to respond is a separate matter entirely, but that has virtually nothing to do with the importance of treating gunfire as an emergency.
For context, my city has a 20 minute average response time for priority #1 911 calls (violent crime in progress). Response time for priory #2 & #3 calls is around 6 hours ( non-violent crime in progress, in progress disputes, just completed felonies, and stolen cars).
There are 10,000 ShotSpotter events a year.
As much as I wish it weren't the case, I simply dont think looking for casings, getting statements, and or performing a search is likely to be the highest demand for police time.
Respectfully, you don't seem to have any experience with law enforcement or military. 2-5 happen routinely.
>There are 10,000 ShotSpotter events a year.
And that is a separate issue of capacity vs demand. Nevertheless, a system like ShotSpotter, when paired with resources and a triage system, is a very real augmentation to law enforcement capacity.
The problems with SpotShotter are completely separate from this. That's precisely why law enforcement seems to want it so badly.
That is the only point that I was making. I already agreed that they would do the things on your list in an ideal world.
My point was that the data integrity is less of an issue for police than their bandwidth to take action on it. Maybe I wasn't clear enough.
Once you are already in a triage situation,the questions is what you cut to spend more time on shotspotter. I think in my city they will sometimes do a drive the area to see if there is anything obvious. They certainly cant ignore a home invasion to collect casings at the scene or go door to door asking questions.
if you're using this tool to dispatch officers to investigate then it's going to lead to a lot of wasted time
The alternative is an extensive investigation which is difficult to determine the location, if not impossible.
I don't mind these shot spotters being installed vs CCTVs as the microphones are really just tuned for gunshots and can only serve one purpose (in the current form)
> If this result is attributable in part to missing or nonmatched records of investigatory stops that did take place as a direct consequence of a ShotSpotter alert, CPD’s record-keeping practices are obstructing a meaningful analysis of the effectiveness of the technology.
Should have been the end of the article right there, but techdirt decided to write one anyway.
And false positives really aren't that big a deal anyway. You get false police reports, at least with the sensors you can recalibrate and adjust.
If only we took companies out to defend their product at face value we could waste even more time and resources!
So much so that it's "obstructing a meaningful analysis of the effectiveness of the technology". Their words, not mine.
> Everywhere it's installed officers find usefulness in triangulating where shots are popped off.
At best, cops are going on 'feels' here because, presumably, the data aren't sufficient to provide a solid basis for analysis. Or the police are withholding data that proves the system works well, which prima facie seems unlikely.
I don't know what the real numbers are, but from that article it seems noone does...
A proper reporting of numbers would be nice.
I would rather the cameras go to the gunshot noises than stay bolted to a pole outside my house, which is where they are now.
Also, a faster response could result in a better rate of finding crime. "89% of ShotSpotter’s reports led police to find no gun-related crime" does not equal "89% of reports weren't gunshots". It could be that police show up too late.
In a sort of magical way where there is a drone that can easily be dispatched from within it's range that has the intelligence to not only navigate to the scene through a complex city environment but also find and gather evidence. I bet any current drone deployment is likely to be slower than dispatching people patrolling and there will be far fewer of them than regular cops. It's basically fantasy when the Shot Spotter stuff doesn't even seem to work.
Drone navigation has been progressing quite a lot over the past decade, and law enforcement in some places is already using them in an automated way, as the article mentions. Also, the suggestion above was forward looking, it wasn't suggesting that the NYPD should glue these systems together today.
It doesn't tell a gunshot from a dumpster lid slamming shut.
It does produce almost entirely false positives.
It doesn't identify the location of a gunshot to any useful level of precision.
I suppose if you want to start a police state, and make every fourth resident a cop, you might be able to mitigate the 90% false-positive rate, and the horrible location-identification rate, but at that point, do you even need shot-spotter?
Right now, police have data about where loud sounds generally come from, and video from fixed poles where crime happened in previous years. Both of these aren't great data points on their own.
The intersection of that data: [Video around locations where loud noises just happened] is probably much better than disjointed data points of [video of where crime happened in previous years] and [general locations of where loud sounds are happening].
If you built towers for microphone above the cityscape you could get an accurate triangulation, but obviously nobody's doing that.
A sensor 4 blocks away will often record a much louder sound than one half a block away, because of buildings that are variously absorbing or reflecting sound.
That makes triangulation utterly useless.
You can localize to maybe a 20 square block area, but that's not a whole lot for cops to go on.
Mind if I ask how a sound recorded by two different sensors can be identified as the same original sound if the loudness is different? Unless the frequency of loud sounds is less than the time taken by the sounds to reach the sensors, wouldn’t it be hard to pick up two sounds spread apart by say a second but know that they are the same sound?
For example, if I’m on the phone with a friend during a thunderstorm and we are on opposite sides of the city, will we hear and be able to identify the same thunderclap?
Does sound have usable distinguishing features other than just loudness?
Than raw, peak amplitude? yes.
Frequency and amplitude variation over time will be roughly consistent in a linear system (i.e. the sound will get louder and quieter over time). If you have a stable, well-qualified listening system (such as, for example, a set of fixed recording locations in an environment that is acoustically static), you can build a map of how sound sources map to each recorder, and use principles of directed arrays to get very accurate location projections.
To use your example, yes it would be difficult for you and a single friend to confirm that you heard the same sound, but you could come up with a probability that if you heard a sound with the same characteristics with a certain time difference and loudness difference, it is likely to have been from a certain location. If, however, you have 4 friends, it gets much less likely that all 4 heard a different sounds calibrated exactly to seem like they came from a specific source when the data is combined.
Back in the shotspotter example, your microphones are in fixed positions, so you can figure out in advance what the echoes are going to be for any given originating sound, and do a reasonable job of triangulation with 3 sensors. You also have some other givens, eg it's reasonable to assume that the location of any actual gunshot is 1-2 meters above ground level; this means you'll be slightly wrong about rooftop snipers/shootouts but those are vanishingly rare compared to street shootings. I don't think anyone claims the product to be reliable for detecting shootings in building interiors, where you'd be competing with televisions and other false positives.
As I mentioned earlier, a product like this can get away with a relatively low success rate, as long as it's slightly better than the information that comes in from 911 calls and the rate of false positives if either tolerably low or negligently unaccounted for.