Columbia Fire Department adopts AI-powered software for improved performance

Erika McGuire
COLUMBIA, Mo. (KMIZ)
The Columbia Fire Department has purchased a software program that uses real-time data analytics to improve overall performance.
Earlier this year, the City of Columbia approved the software, known as Darkhorse Emergency, at an annual cost of approximately $44,000.
During a candidate forum last year, Columbia Fire Department Chief Brian Schaeffer expressed his vision of wanting Columbia to be at the forefront of technology, using artificial intelligence. That vision became a reality when the fire department started using the system in July.
“Darkhorse is a software program that integrates with our computer-aided dispatch and our records management system to take all of the information that’s out there in our system,” Schaeffer said, “inclusive of our calls of service, over the last five years and all the parameters: things like how long it took us to get there, or what the time of the day was, what the traffic, slowdowns were occurring during that specific time.”
Additional data, such as road speeds, road complications, speed bumps and system occurrences, are fed into the program, allowing the fire department to ask questions like where additional staff may be needed or where a new fire station should be built.
The program is monitored by the fire department on a minute-to-minute basis, Schaeffer said, and it helps make decisions for the future.
“Recently, with Station 5 and Station 10 and probably future Station 11, Station 12, as we continue to build, tax dollars are finite,” Schaeffer said. “Every dollar that we invest, we want to make sure that investment works for today, but it also works for 20 years from today.”
CFD is required to get an “A” rating or a 90% when responding to every single call, Schaeffer said. By using the Darkhorse program, the department can predict how changes, such as relocating fire trucks or increasing staffing at a station, will impact their overall rating. This allows for data-driven decisions that help maintain or improve performance standards.
Schaeffer said the department wants to hit its standards 90% of the time, and extra resources can help meet those goals in some parts of town.
Darkhorse Emergency can also highlight risk areas throughout Columbia.
The software looks at modifiers such as the low-income population, the population that is 65 and older, people with a disability and the number of children under 5.
Schaeffer used University Hospital as an example. It has a “very high risk” of needing CFD’s services — in the 97th percentile, Schaefer says.
Probabities are based on past patterns, he said.
The program also uses AI-generated projections for up to five years to determine the most effective locations for fire stations, based on current jurisdictional boundaries. The department can run scenarios like not having any existing fire stations in Columbia. The program will then identify locations for new stations using parameters provided by the fire department.
“The station [Station 2] would move a little bit to the south,” Schaeffer said.” But interestingly, Station 5 is in the exact same place. So when we looked at moving Station 5 because it’s 50 years old, it needed to be replaced.” We ran the Greenfield and lo and behold, we are in the right place based on current and future growth.”
Since beginning to use the Darkhorse software in July, Schaeffer says it has already made a difference at CFD.
“We’ve made assumptions on the performance of the system, and really, what we’re capable of, meeting our goals and establishing some levels of excellence that before we had tucked away as, ‘One day we’ll get to this,'” Schaeffer said. “Well, now we’ve got the ability to not only get to this, but model it and then take that information that we garner from that model or that scenario and put it on the street.”