Stopped at the door

Let's say it's New Year's Eve. And after a late night partying across town, you somehow make it home. Cold and tired, you take out your keys - but the door is already open. Someone has forced their way inside your home. Laptop, sound system, TV, microwave: all gone.

Being the victim of a break-in is every tenant's and homeowner's nightmare.

The chances of ever seeing your belongings again are slim to none.

So the best thing to do is prevent the crime in the first place. But how?

In Germany, police and other security authorities are trying to get better at prevention, using some of the latest software.

Following crime patterns

One such piece of software is developed by the Institute for Pattern-Based Prediction Technique (IfmPt) in Oberhausen.

The program is called Precobs: Pre-Crime Observation System.

It collects data from the most recent and past break-ins and analyzes them for patterns.

Police hope it will help them identify high-risk areas.

"The program can develop focus areas with the data and help us understand burglary patterns," says Michael Schweer of IfmPt.

Four main factors are used in the analysis: crime scene, time of crime, the things that were stolen, and the thieves' course of action.

The basis for this software is a concept called "near repeat." It says that if there's been a burglary in an area, there is a good chance of a repeat offense in the same area within 72 hours.

Police can also tell which seasons see the most break-ins - New Year's Eve is a good example of that.

But beyond that, analysis is difficult because of the volume of data.

And that is where the software comes in.

"The problem is the patterns are usually rather complex," says Schweer. "The software can analyze tens of thousands of data sets in mere minutes. A human could never do that in such a short time."

Tough for professional burglars

So instead of wasting personnel on this work-intensive analysis, the software would only require a few officers to operate it.

They could then decide on location how to "react to each prognosis and which preventative measures to take," Schweer says.

One measure could be sending plainclothes police officers to high-risk areas.

Schweer says he does not believe burglars will react to the software by changing their strategies.

"Even if the criminals adapted their behavior, you could still see a pattern," he says. "We focus on repeat offenders, professional burglars and gangs. And they are definitely going to follow certain patterns."

The German states of Bavaria and North-Rhine Westphalia are keen to test the systems. Police in Cologne and Duisburg will use prediction software from early 2015 until September 2016.

The results will then be examined to see whether the software truly delivers.