In a reality-meets-sci-fi story, the New York Times reports today that several cities across the country are using computer programs to predict crime hotspots and send police to nab crook before anyone has broken the law.
The arrests were routine. Two women were taken into custody after they were discovered peering into cars in a downtown parking garage in Santa Cruz, Calif. One woman was found to have outstanding warrants; the other was carrying illegal drugs.
But the presence of the police officers in the garage that Friday afternoon in July was anything but ordinary: They were directed to the parking structure by a computer program that had predicted that car burglaries were especially likely there that day.
Sure it’s like the movie Minority Report but without the precogs. A computer program takes previous crime data, analyzes current outbreaks and with a little pixie dust, predicts where crimes are likely to happen at a future time.
Santa Cruz’s method is more sophisticated than most. Based on models for predicting aftershocks from earthquakes, it generates projections about which areas and windows of time are at highest risk for future crimes by analyzing and detecting patterns in years of past crime data. The projections are recalibrated daily, as new crimes occur and updated data is fed into the program.
On the day the women were arrested, for example, the program identified the approximately one-square-block area where the parking garage is situated as one of the highest-risk locations for car burglaries.
So my question is: how much of this “science” was derived from military analysis of insurgent networks and, on the flip side, how can something like this be useful in the COIN and CT environment? Seems to me it could work well with predicting IED emplacement, especially since we have about 10 years of data in Afghanistan to feed into a predictive and anticipatory model.
In this case, a tool like this could be used to emplace a sniper team over a route or run clearance in one area instead of another. In a COIN environment, we might be able to use this to predict insurgent “shadow government” takeovers and strong-arming, or predict infiltration routes from Pakistan and times of ingress.
Very intriguing … and we don’t need to maintain a team hairless psychics in pink ooze to do it.
{ 13 comments… read them below or add one }
Our EMS has a similar program that predicts vehicle accidents and shows you hotspots where you could place an ambulance to shorten response times. Regression analysis can help you quantify all sorts of spatial data which I'm sure is already being used in theater. I'm sure if EMS in podunk Texas has it, NSA and NGA have access to even cooler stuff.
Just wait for BIG Brother to arrest people for crimes they didnt do or had intentions to do and go after more innocent people.
You do realize in the US you can't be held for more than 24 hours without being charged of a crime? You can be stopped for suspicious activity, but unless you have a warrant out for you, or (like in this case) have illegal drugs or even stolen objects you can't be arrested? So, even if the computer pulled you're name up and said "this guy is going to murder someone tonight" they can't arrest you, until right up until you at least attempt the murder.
You can live in your paranoid, delusional world all you want, but I'm happy right here in reality.
Until we suspend those rights in the name of safety.
Cause that’s how democracy works… Glad I don’t live in your world, Jim
Yeah Matt, that's how it works, ever hear of TSA? DHS?
Cool. Sophisticated criminals might incorporate a degree of randomness into their crimes. Pick 6 targets then roll some dice to choose which one. Or flip coins to determine the day for the crime. Luckily most criminals are not sophisticated.
Ubiquitous cameras and behavior analysis would also do this. Anybody who is looking through the windows of multiple cars is probably up to no good, cameras could spot that and direct the cops to ask the person a few quick questions.
The Panopticon is inevitable. It will reduce crime. It will also be abused at times. But people will welcome it. I used to live not far from some low-income housing projects that produced a lot of criminals. I wish my apartment complex had had more cameras cause the criminals occasionally wandered over there. The local grocery store there went out of business cause the thugs stole so much and hung out in front, scaring off the paying customers.
By choosing 6 targets, you'd already lose the randomness.
But the most common type of die is a 6 sided die which goes into JTG's thing about rolling dice to pick your target. I suppose that if the criminal is a gamer they could randomly choose anywhere from a 4 sided all the way up to a 20 sided dice to roll in order to choose their target; random enough for you?
This sort of thing has been used by us marketing folks in the business world for years. The nice thing about applying it to large scale things like COIN and crime is that the more data you have the better the results. Predictive analytics can seem a little out there at first but it's getting very, very good to the point that in a few years in will be quite disconcerting. While adding some randomness can delay accurate results you would be surprised how quickly a new pattern emerges, especially when you've got some very good computers crunching new numbers from the daily batch run of information.
I'm against the ever present cameras as they aren't nearly as successful as more folks on the street making more notes and feeding such into a competent system. I've yet to see a winning cost benefit analysis on the expense of a camera system. In most cases someone's finger is on the scale, much like the traffic cameras that are now being caught out.
As for folks loitering and killing businesses here are two successful suggestions. As we age our ears can no longer hear certain frequencies, but the younger folks can. You can now buy sound recordings that are quite unpleasant to teenage ears but that cannot be heard by older folks. On for a more low tech approach, play classical music. Marketing studies showed younger folks self selected and left those areas alone while it attracted older patrons with higher purchasing power.
Ohhh! now I understand. A previous post made this seem more science fictiony and likely to infringe upon our rights but this is actually a legitimate way of anticiapting crime and reducing response times. The cops weren't arresting people based on a prediction they were put in place to prevent a crime from occurring and arrested someone who was already in need of punishment. Sounds good to me.
Not really finding bad guys before they've been bad, more like an exercise in statistics/probability.
Finding the next unabomber or Martin Bryant with technology before they've committed anything, now THATs an impressive feat
In response to the article above they are already using a program very simalier to spot IEDs and the terrorists that plant them, the article came out in stratigypage about a year ago if I remember right.