FAAC Commercial


Data-Driven Fire Simulators and Models: Better Data Makes Better Firefighters

Data-Driven Fire Simulators and Models: Better Data Makes Better Firefighters


Fire simulators and fire-fighter training have not been left out of the age of “Big Data.” For example, the City of Atlanta Fire Rescue Department has worked with analysts from the Data Science for Social Good project to streamline their inspection process. This ultimately resulted in Firebird, an open-source framework that uses algorithms to determine which attributes of a commercial structure are the best predictors of fire risk. It then identifies commercial properties missing from inspection lists, and assigns them an inspection priority based on fire risk (as determined through data analysis). It displays all of this on an interactive map so that inspectors can best prioritize inspections and personnel allocation. The Los Angeles Fire Department uses a similar tool (WIFIRE) to both monitor existing wildfires and predict their path.

Other tools push beyond efficiently organizing what we know, to reveal previously invisible patterns in fire safety. For example, the New York City Fire Department (FDNY) relies on a system similar to Atlanta’s Firebird, called FireCast, to help prioritize property inspections. FireCast ranks structures along 60 dimensions related to structural fire risk. In correlating so many dimensions, FireCast surfaced some interesting findings. For example, FireCast found that active tax liens on a building and ongoing foreclosure proceedings are both strong indicators of increased fire risk.

In 2015 the City of New Orleans and the New Orleans Fire Department launched an Analytics-Informed Smoke Alarm Outreach Program. This program leverages data analytics to conduct a “targeted, risk-informed door-to-door smoke alarm outreach program.” This allows them to use strained resources more productively, prioritizing residences that are both least likely to have smoke alarms and most likely to see fire fatalities.

Integrating Data Analysis into Fire Simulators and Firefighter Training

Predictive models aren’t the only ways that new technology and data collection systems are helping fire brigades improve their performance.

Phil Duczyminski is a fire training officer for the City of Novi, Michigan. He’s found that one of the key challenges they face is identifying very small mistakes trainees are making early on before these become bad habits in the field.

“A lot of times,” Duczyminski notes, “those small mistakes will snowball into large mistakes” on the scene. That can prove deadly.

Fire simulators with the proper built-in assessment tools help trainers pinpoint the little mistakes most likely to snowball in the field. This is part of the reason that Duczyminski favors FAAC fire simulator solutions. FAAC training sims are built around active student assessment systems that include After Action Scenario Review (AASR) and student performance assessment and report functions.

“Just about the majority of firefighters are tactical/practical learners,” Duczyminski points out. These are individuals who learn best with their hands and need to see something to believe it. Systems like VITALS (the Virtual Instructor Trainee Assessment Learning System) track every decision the trainee makes, and every interaction they have with every switch, lever, or control element. It also integrates a full audio and video capture of the entire training evolution. Having all of this data on hand for consideration after the exercise gives trainees a powerful perspective on how they handled a scenario. “When you can point to the screen and say, ‘See that right there? See how you’re doing this?’—that’s a very, very powerful teaching tool.”