You’re strolling along Acacia Avenue just like last week and the week before: the rucksack slung over your back is for the swag. This is a well-off area: middle class mansions in treed gardens with plenty of cover – they do like their privacy here. Your preference is for jewelry, cash, laptops – light and high-value stuff that’s easy to walk off with.
It’s a sunny afternoon, mid-week, and as usual everyone’s at work. The last two times it was a walk in the park: amble up the drive as if delivering something, knock at the door and if no response then dodge round the back and force the window.
You pay no attention to the white van parked at the sidewalk and you’re oblivious to the small box-shaped devices which have appeared high up on the lamp posts. If you bothered to look up, you’d probably think the quiet shape wheeling above was a bird.
As you had hoped, your first choice of house is ideal. You walk up the drive, lined as it is with helpfully-obscuring bushes. There’s no-one home so it’s into the back yard, pull out your hardened-steel crowbar and in a second lever open a window. There’s no real need to rush but ten minutes later, your bag satisfying full, you’re back on the sidewalk.
And all these people in blue uniforms with guns pointing at you – where on earth did they come from???
Put aside the tools of close-in tactical surveillance: the white van with its camera crew, the WiFi CCTV cameras on the street furniture, the police drone tracking your every move; how did the police know you’d be there at all? The answer is that they were sent there by a predictive policing program, based on UCLA software which normally predicts earthquake after-shocks.
Here’s the future of predictive law enforcement. An automated crime forecast (like a local weather forecast) is sent to each area commander who then makes force-deployment decisions. Police officers receive detailed updates on crimes about to be committed along with details of criminals in the locality. The predictive computer system is fed real-time information from crime scenes, witness interviews and the justice system which drives the predictive-analytics program. It works because criminals are both copycats and creatures of habit.
Somewhat surprisingly the police are not completely won to this idea. Some complain that their intuitions and ‘street smarts’ are being ignored in favor of a mechanistic and remote set of algorithms.
“There is the science of policing, and there is the art of policing,” said LAPD Deputy Chief Michael Downing, who relies heavily on technology as the head of the department’s counterterrorism efforts but remains wary of predictive policing. “It is really important that we learn how to blend the two. If it’s all about the science, I worry we’ll lose the important nuances.”
Lt. Sean Malinowski, who oversees the LAPD’s crime analysis unit, wants to head off civil rights concerns about profiling and harassing individuals before they do anything wrong. He denies that the technology is a real-life version of “Minority Report,” a 2002 science fiction film in which cops arrest people for crimes they are about to commit.
But if predictive policing turns out to be a success then expect further advances. There are already plans to build a far more sophisticated model to replicate actual buildings in real neighborhoods in Los Angeles which could be populated with computerized criminals and used to predict the movements of actual law-breakers. If this were combined with the GPS-tagging and automated video surveillance of known criminals out on parole (or for a period after release) then the impact on the crime rate might be very significant indeed.