Friday, November 18, 2011

Predictive firefighting

I've blogged before about the concept of predictive policing: using our data and our knowledge to make informed judgement about where crime is likely to occur in the future, what kinds of crimes are ascendant, and who is likely to commit crime.  Google "predictive policing" and you won't go far until you find my name associated with this topic.  Earlier this year, I edited an issue of Geography and Public Safety on predictive policing, and contributed my own article to the volume, "Police Legitimacy and Predictive Policing."

This week, I participated in an intriguing meeting and web conference, along with Lincoln Fire Chief John Huff, Assistant Chief Rick Furasek, Chief of Logistics Kendall Warnock, Chief of Training Roger Bonin, Capt. Eric Jones, and Capt. Scott Weibe. The web conference was with Buxton, a firm that specializes in consumer analytics.  Buxton has been doing some work with the Philadelphia Fire Department, applying its methods to create a "Fire Vulnerability Index."  This all sounds incredibly familiar to me, and is essentially the same process used in policing to deploy resources and determine strategies that are geared to creating the most bang for the buck.

Our current station relocation study is well on its way to identifying the best alternatives for providing data-driven decisions for Lincoln's fire, rescue, and emergency medical services.  While the current study is based largely on historical data about incident trends, the methods used by the private sector for locating retail businesses hold promise in creating more robust predictive analytics in public safety.

5 comments:

Anonymous said...

Director,
Is Arson a big cause of fires historically in Lincoln? I was wondering if you had a breakdown of the types of fires you respond to: residential, vehicle, industrial, dormitory, Hotel, Motel, Bars, restaurants, retail, convenience stores etc? Do you map out other demographics like age, neighborhoods, rental property or homeowners property? I would think other than Arson, fires are unpredictable events but are there patterns?
Just curious.

Gun Nut

Steve said...

Have you ever thought about using all these state-of-the-art data gathering, analysis, and prediction methods to cash in at the horse races? :)

Anonymous said...

Anyhow, guess what?

Anonymous said...

PSD,

This isn't related to the post topic, but relevant to public safety. I saw that Omaha, a little less than twice Lincoln's pop, has had 33 murders in 2011 YTD. Some say it's only 30, because 3 of those should be hidden in the 2010 stats, when the murders were committed, not when they were confirmed.

So, Lincoln, less than an hour away by car, should have had 17 or 18 murders so far this year, but how many have we actually had? I bet you have that number in your head, without having to look it up. Thanks in advance, for any response.

Tom Casady said...

8:57,

Four in Lincoln so far in 2011.

Gun nut:

We are working on developing capability for such analyzes now. We have such a small number of structure fires in Lincoln that we really need to pay attention to research literature that has analyzed fire risk factors in much larger data sets.