Reading Police Attribute Drop in Burglaries to New Strategies
Taken from this WFMZ article reported on December 17, 2014.
Software helps Reading Police Department predict time, place crime is most likely to occur
READING, Pa. – Despite a double-digit drop in the number of officers on the force, Reading police have seen a double-digit drop in burglaries throughout the city. New policing strategies implemented by the RPD over the last year have contributed to a 23 percent drop in burglaries, according to Chief Bill Heim. The department, Heim said, has combined predictive policing data with its existing crime-mapping system, allowing it to shift officers’ efforts from places that have experience crime to areas where the next crime was predicted to occur.
“With almost 100,000 residents, we need to be doing everything we can to keep Reading residents safer,” Heim said. “We’ve lost over 45 officer positions due to financial constraints but have gained this invaluable tool.”
History of PredPol in Reading
The RPD began using predictive policing software, called PredPol, in October 2013 to help it identify where to strategically deploy officers in the places and at the times crime is most likely to occur.
Using a unique algorithm, PredPol provides Reading crime analysts and officers with customized “boxes” on maps of the city, identifying the 500-by-500-foot areas where the highest probability exists for a crime to occur, explained Heim, adding that the boxes are automatically generated for each shift of each day.
“We’ve been tracking the number of times our officers patrol inside the PredPol boxes. Their presence really helps deter crime,” Heim remarked. “It’s shown itself to be a real force-multiplier in crime prevention.”
How PredPol is Used in Reading
The chief noted that some of the boxes represent places that officers routinely cover, but many others are places that might not otherwise receive attention. PredPol’s algorithm is based on three data points: Place, time and dates of past crimes, said Heim, noting that it does not include individual or demographic information.