I predict a break-in: Yorkshire police use cutting-edge technology to deter burglars

A Yorkshire police force is aiming to cut crime using pioneering predictive technology which shows officers specific streets where burglars are likely to strike.

West Yorkshire Police are piloting a predictive policing algorithm designed by University College London academics.

University College London (UCL) academics have created the algorithm for West Yorkshire Police which predicts with greater precision where burglaries are likely to take place so the force can decide where to send its patrol teams.

Police in West Yorkshire, which has had one of the highest burglary rates in the country for a number of years, are piloting the technology in Bradford East and say the number of house burglaries has already fallen.

Sign up to our daily newsletter

The i newsletter cut through the noise

It has previously been using a system which predicts burglary hotspots on the basis of much larger areas, known as ‘quadrants’ or ‘cells’, whereas the UCL algorithm predicts which areas are vulnerable to criminals on a street-by-street level.

West Yorkshire has had one of the highest rates of burglary in the country in recent years.

Unlike Kent Police, which has a contract with a US technology firm PredPol to use similar technology, West Yorkshire Police do not pay UCL for the algorithm as it is part of a collaborative research project.

The UCL team used West Yorkshire crime data and details of its street network to create the algorith, a set of rules which help generate suggested patrol routes for officers to follow on each shift, based on where crime has taken place in the previous days.

It is hoped that, at a time when police resources are stretched, this will allow fewer officers to be on specific ‘hot’ streets for less time but still have a deterrent effect on burglars.

The algorithm has been in use in Bradford East since December 2016 as part of the Patrolwise scheme, which also includes officers being given new mobile devices where the patrol plans for each shift are sent.

Police in Bradford get details of the suggested patrol routes on new mobile devices, designed to help them spend more time on the street.

Dr Toby Davies, a lecturer in the Department of Security and Crime Science at University College London, said he and his colleagues had worked on a number of predictive policing projects in recent years.

They were approached by West Yorkshire Police and asked if they could produce a tool that could be used by the force.

Dr Davies said: “Doing things at a street level rather than at cell level is, as far I know, unique in that regard. There are things that we are doing that you won’t find in commercial products.”

He added: “This is a research project for us and we are trying to see whether this kind of thing works in West Yorkshire.

West Yorkshire has had one of the highest rates of burglary in the country in recent years.

“The bigger question for us is how many patrols you need to have an effect. The motivation for us to do this kind of thing is to maximise the use of our resources.”

West Yorkshire Police has historically had one of the highest burglary rates in the country, and in the 12 months to March had the second highest number of domestic burglaries per head, behind Greater Manchester.

The force has been using predictive policing techniques to try and reduce its burglary rates for a number of years, and in 2012 launched a scheme called Operation Optimal in Leeds, which highlighted wider areas where burglars were likely to strike.

Dawn Wilkinson, project manager for the Patrolwise scheme, said that since the project began in December 2016, Bradford East had seen burglary rates fall in comparison with the same year and with the rest of Bradford district.

Police in Bradford get details of the suggested patrol routes on new mobile devices, designed to help them spend more time on the street.

She said: “While there will also be other factors that contribute to that reduction, Patrol-Wise has given officers an extra tool in their crime prevention tool kit.”

She added: “If we can work it the way we want to, by putting in fewer resources for less time, it is looking positive.

“We are still in the pilot phase. It looks as though the algorithm can help our crime reduction as well as problem-solving and doing other things. We can also use this for other crimes such as theft from motor vehicles.

“We are still testing the algorithm and working with UCL, but the project for me is going well.

“We are not saying ‘this is fantastic, we are going to charge away with it’, we are testing it as much as we can.”

Predictive policing is partly based on the idea that burglars are likely to target the same streets repeatedly once they have struck once, as they are more familiar with the area.

Police suspect they often return to the same streets once their victims have replaced the stolen goods, prompted by seeing boxes containing new TVs and radios outside their homes.

Mrs Wilkinson said: “Burglars don’t want to get caught so the best way to deter them is with yellow jackets and PCSOs. If the burglar sees that he is going to go away again.”

This summer, The Yorkshire Post revealed that West Yorkshire Police was planning to ‘rebuild’ its neighbourhood policing after the loss of hundreds of officers led to criticism that the force had lost touch with its local communities.

Bosses blamed funding cuts and an increase in demand for its decision to dramatically reduce the number of officers carrying out problem-solving and engagement work on the streets of the county.