Friday - May 07, 2021

Can Police Foretell Crime?


Law fighting agencies world wide have  been losing the battle in the technology war. However, this is about to change as many urban police departments are embracing modern technology to thwart criminal efforts before they happen.

The Los Angeles Police Department (LAPD) is leading the way to computerize the fight against sophisticated criminal gangs in this digital era. If you have been to downtown LA, you may have noticed a number of criminal analysts and technologists sitting in front of video walls extending up to 10 meters wide—they are not Microsoft tech wizards who have paid the department a courtesy call; they are police officers at work. This is the Real-time Analysis and Critical Response Division of the LAPD and they are  solving crimes before they happen.

It is not in LA alone, a dozen other police departments around the world are preventing robberies, murders, rapes and kidnappings before they take place. And all this has been possible through predictive policing.

How the police are predicting and preventing crimes

It all started with one man’s effort dedicated to help the police force. P. Jeffrey Brantingham, an anthropology professor at UCLA, developed a predictive police program that is now popularly known as PredPol. He sees it as the end game to random crimes in our cities. The program is based on a science fiction report known as minority report which is about prediction on when and where a crime is likely to take place. Police can then piece up this information and prevent the crime from happening. So far, it is LAPD that is thought to have busted the most robberies, burglaries and other forms of crimes since they started using PredPol.

How PredPol predicts the crimes

predpol futureThis program is based on algorithms collectively called predictive policing. The algorithm uses data on past crimes and analyzes it to point out areas which are likely to be hit by a crime. Once it identifies those areas, it places little red boxes on those areas in the map of the city or the region covered. Criminals tend to be territorial minded and will keep coming to a place they had successful loots. Police on patrol are then alerted of the red spots. They then lay low around those areas until they see the criminals coming in to execute their plot. Using this prior espionage, the police are able to arrest or at least prevent the crime from happening.

John Romero, Real-time Analysis and Critical Response Division’s top official, draws a lot of similarity between this process and fishing. An amateur fisherman may not be good enough to predict where fish are but a good fisherman is able to go where most fish are. Good police officers are therefore able to piece up the intelligence from PredPol and know the exact locations where crimes will take place.

PredPol is the new high-tech tool police are using to collect, store and analyze data faster and cheaper. This has enabled police officers to be fore-informed and prepare to prevent crimes in real-time.

However, this is not something that began recently, it's been a progressive effort that is now bearing fruit. For instance, the New York Police Department started statistical crime predictions in the mid 90s. Back then, they were able to make a number of arrests and prevent crimes before they took place. Since then, many global police departments have invested in data intelligence efforts that are aimed at predicting and keeping abreast of criminal trends.

Challenges facing predictive policing

Predictive policing may be the cutting edge tool for police departments, but it also comes with challenges. The first challenge is the historical dullness of merging policing with technology. John Eck, criminology professor at the University of Cincinnati, says there is a lot of monotony and unnecessary police questioning as a result of the red boxes in crime hotspots. He says that if the predictions keep pointing to the same areas, then the operators of the targeted areas have something fundamentally wrong with their methods. There is no need to bring in the police and disturb the peace of a whole precinct  when the problem can be dealt with privately.

The second challenge is that the police departments (especially in the USA) have gone a notch higher and added drones to the predictive policing. LAPD for instance, has 1000 or more closed circuit cameras feeding their drones and recording real-time activity of all citizens even those who are not criminals.

The FBI are planning face-recognition

face recognitionThe FBI is planning to take their efforts a notch higher by the launch of face-recognition technology that will contain at least 52 million record entries. A lot of non-criminal related data will be collected for no good reason. This is why the Electronic Frontier Foundation (EFF) has  filed a lawsuit against FBI. EFF argues that the police agencies have unlimited access to records of innocent citizens. This is not safe as witnessed in the case of the Californian woman who was held at gunpoint by police for mistaken identity. The police database had wrongly identified the woman’s car as stolen. She later sued the police and won the case. More of such cases will keep coming up and the compensation payouts are predicted to be massive.

Take the Edward Snowden example; he leaked documents secured from civil surveillance by the US National Security Agency.  This did immense  damage to the push to embrace  predictive policing. The public is now aware of privacy breaches, and we will see more court dramas when PredPol and other policing software comes to aid of police work.


It is not a bad thing to have police officers foil crimes before they happen. If only it can be done within the corridors of privacy then this will be a great step towards a crime-free world. There is no need for non-criminal data to be captured by the law enforcement agencies without the knowledge of the citizens. Until the challenges of predictive policing are fully dealt with, overwhelming litigation between governments and their citizens will be common.