Predicting crime with 21st century tools and experience

Ronald Fellows is IBM's Smarter Public Safety Analytics Global Lead.

Law enforcement has been built on a core tenant: Bad guys break the law and good guys try to catch them. It’s a pretty basic, but effective approach especially when you consider that law enforcement agencies continue to see reductions in crime. The challenge is that this approach continues to prove that it’s time consuming and expensive.

Just look at the average crime, officers’ start with little to no information. They are essentially starting from scratch with each report that lands on their desks. Now they have to generate reports, gather statements, chase leads, conduct follow-up meetings and often rely on their experience make certain assumptions. And if they catch a case at the end of their shift, it’s almost guaranteed that they are going to have to put in overtime -- a stress on them and the department’s budget.

When they do have a suspect, this creates more paperwork, phone calls, evidence to sort through and interfacing with the judicial system. This isn’t like an episode of the cop shows where everything is solved and the bad guy is behind bars in 60 minutes or less. Solving a real crime takes weeks and sometimes even months. All the while, the criminal can cause more havoc on the community.

I’ve oversimplified the process.

But what if law enforcement agencies were able to prevent crime all together? Well, maybe not all together, but put a significant dent in preventing crime … in other words, predictive policing.

Predictive policing is a concept that has been emerging over the course of the last few years that brings together crime analytics tools, crime fighting technology, intelligence-lead policing, etc. to provide agencies with the insights they need to develop strategies and tactics to reduce crime. It’s about centralizing what they already know about crime and coupling that with other information (including an officer’s experience and knowledge) and analyzing it to be able to “predict” where and when crime is likely to occur.

Here are just a few benefits to this approach:

• The ability to dynamically schedule patrols and dynamically reroute patrols to maximize public safety
• Better scheduling of officer availability and manpower planning
• Better planning of equipment availability (such as vehicle maintenance)
• Better training and skills development
• Better management of deployment and recruitment
• More efficient use of available resources
• Financial savings

Businesses have been using predictive analytics for years to anticipate market trends, spending habits of consumers, credit scoring and even changes in the weather because all of the impact it has on their bottom line. Now police departments are taking a page for big business’ playbook. In New York, Los Angeles, Memphis and a growing number of other police departments have already begun to deploy this model and have seen sizable reductions in crime. Many reduced crime by nearly a third and have been able to better utilize their existing recourses so the sting of the economic downturn didn’t impact them as much as it would have if they had stuck to traditional policing.

While I’m the first to admit that predictive policing isn’t the singular silver bullet to eradicate crime from our society, I do firmly believe that it will allow the people who protect us on a daily basis with an invaluable new weapon in their arsenal.

Photo: The Charleston Police Department is working with IBM to assist the city’s more than 400 police officers, such as officer Joseph Dela Rose, to more accurately evaluate and forecast crime patterns. (Credit: Jim Mossman)

This post was originally published on Smartplanet.com