< Browse more articles
Fighting Crime with Big Data

Imagine how easy the job of law enforcement would be if only we could predict exactly where and when crimes were going to occur. Well, even though predicting the future is still largely out of the scope of human capability, technology is advancing in such a way that we are able to hone in on the locations, times of day, and types of criminal activity that are most likely to occur in certain neighborhoods.

Called predictive policing, the technology refers to the use of mathematical, predictive and analytical techniques to aid law enforcement in identifying potential criminal activity. In other words, it’s not a crystal ball that foretells the future precisely, but an effective tool to help minimize crime, in areas and at times when they are projected to occur most.

Since 2011, Predictive Policing techniques have been readily introduced and implemented by police departments in jurisdictions south of the border—including California, Washington, South Carolina, Arizona, Tennessee, and Illinois—and now the technology is catching up here in Canada, too.

And naturally, throughout the past few years, this technology has only gotten better. For instance, Motorola Solutions’ CommandCentral Predictive, a state-of the-art application that helps agencies to more effectively direct their resources and time with intelligent, targeted-area crime predictions, has proven to be three times as accurate as traditional hotspot analysis.

Predictive analytics vs. hotspotting

Predictive analytics is more than just hotspotting in that it provides daily, focused crime forecasts that predict what crimes will occur and when, along with tactical details so that officers can act on the information in real time, while out in the field.

Using your police department’s Records Management System (RMS) data, the software learns from historic data, specific to your jurisdiction, and creates a prediction model for future crime. Officers receive detailed information on each prediction area—field tested to accurately predict, on average, 30% of next-day crime—while also obtaining full narratives and arming officers with everything they need to take action.

Actionable predictions, proactive response

Imagine that a rash of robberies has taken over a community over the course of one year. CommandCentral Predictive identifies that those robberies typically occur around 4 p.m. in specific areas. A crime analyst sees this information visualized on a map and identifies similar narratives, all mentioning a white Ford vehicle for the group of robberies. Using the data, CommandCentral Predictive is able to project where the next robbery is most likely to occur, and when. The analyst then recommends directed patrol at those times. Officers in the field see the same information as the analyst, in addition to past arrest records of offenders. The officers use that information to focus on key suspects, leading to an arrest during a robbery.

Data integration and automation

By integrating with various data sources, analytics and automation are applied to provide actionable intelligence to personnel across public safety operations. So, whether it's in the command centre or officers out on the street, they’ll now have capabilities to improve strategic and tactical planning, incident response and investigations for safer, more effective policing.

Key features of CommandCentral Predictive
  • Targeted area predictions created by day, area, crime type and shift, so they are relevant for each officer
  • Detailed prediction information so officers know where, who and what to look for—including relevant historical incidents and narratives
  • A data quality and validation tool, eliminating common entry errors
  • Device agnostic, allowing for seamless work from desktop to mobile
  • Street views for quick, easy reference and greater situational awareness

For more information about Motorola Solutions’ and Predictive Policing, please visit www.motorolasolutions.com/predictive.

M Sltns blu rgb h pos