IBM teams up with police department to predict 'hot spots' for crime

Monday, January 30, 2012 - 2:26pm

What if there was a way to better stop crime? New technology by IBM is helping police in Rochester, Minnesota do just that.

Every day officers are responding to several different scenarios, so with the information police already have the software can: identify patterns, link criminals who may be working together, as well as forecast crime "hot spots."

The software is called IBM InfoSphere Identity Insight.

The decision to purchase the program stems from the police department's move back in 2010 to take on an intelligence-led policing philosophy.

"This particular type of software is more adept at identifying who's who, who's doing what. So we can target our most prolific and serious offenders as opposed to more of a shotgun approach," Lt. Timothy Heroff said.

The product wasn't developed at the IBM facility in Rochester, but software developer Mike Koranda works on similar advanced analytics software used for businesses.

"So we build general tools that can be applied to many industries and domains. The law enforcement area is just one area that they can be applied to and they would be customized by use in that specific area by people who are the domain experts," Koranda said.

Although the project is still in its early stages in Rochester, the police department hopes it reflects what's been seen in other cities that have used the software, often times experiencing a decrease in crime.

"This approach is very different in that we're using data to try to prevent crime in the first place," Lt. Heroff said.

Since Memphis started using analytics technology crime has been reduced by 30%.

Lt. Heroff also says it will increase safety for police officers, giving them a better idea on who they're up against when responding to a dispute.

The software was ordered Wednesday.

It's expected to be implemented within the next month.

The police department should start seeing results within six months.

Comments News Comments

Post new Comment