New Analytics Tool Helps Predict Where There’s a Lack of Home Smoke Detectors

small1 /


Connecting state and local government leaders

Smoke Signals is open source, free and uses publicly available data.

In an effort to help reduce the number of lives lost in home fires each year, a tech company launched an online analytics tool on Thursday, which aims to map the locations in dozens of U.S. cities where households are least likely to have smoke alarms installed.

Getting the alarms into more residences is seen as a key way to prevent fatalities in fires.

Designed by the New York based company Enigma, the tool is called Smoke Signals. It’s open source, free and uses publicly available data. Currently, it includes interactive maps featuring granular geographic areas known as census blocks for 178 U.S. cities. Each blocks is scored based on the estimated risk that homes there do not have smoke detectors in place.

The scores are based on a predictive statistical model that incorporates data from the U.S. Census Bureau.

“Our hope is that it can be adopted by major—by any—U.S. fire departments,” Marc DaCosta, Enigma’s chairman and co-founder said during a phone interview on Thursday. “This is really intended to be a tool for someone who is trying to organize on-the-ground outreach efforts.”

Noting that 25,000 people are either hurt or killed in fires around the country in a typical year, he added: “One should be able to drive that number down.”

According to the National Fire Protection Association, between 2007 and 2011, 37 percent of U.S. home fire fatalities happened in residences without the devices.

National Fire Incident Reporting System data for 328,346 residential blazes between 2011 and 2013 show that at least 71,957 of the incidents took place in homes without smoke alarms. During that same timeframe, an average of 2,530 people died each year in residential fires.

Cities, such as Washington, D.C., and Seattle, as well as organizations like the American Red Cross, have programs that provide free smoke alarms to qualified residents.

But it can be hard for those doing outreach to know where to find the people that lack the alarms—which doors to knock on in other words.

“The conventional wisdom of how to solve that problem was just to, kind of, go on a gut feeling, or go set up a table at the mall,” DaCosta said.

That’s where Smoke Signals comes in. Instead of an ad hoc approach, outreach efforts can be directed toward households that lack the devices.

The analytics tool has roots in an initiative Enigma worked on in New Orleans earlier this year.

A tragic fire struck struck a residence without a smoke alarm in the city’s Broadmoor neighborhood in 2014, killing five people, including three children. The incident drew attention to the problem of homes without smoke detectors. The five deaths were among 22 fatalities that occurred in structure fires in New Orleans between 2010 and 2014 and, as Mayor Mitch Landrieu’s office noted in March, “In nearly all cases, no smoke alarms were present.”

So the city decided to develop a predictive analytics model that could guide door-to-door outreach efforts meant to get the alarms installed in more city homes.

Enigma provided technical assistance and peer review on the project.

Nobody at Landrieu’s office could be reached for comment on Thursday about the status of the initiative.

But, for Enigma, the work in New Orleans set the stage for Smoke Signals. “We already had the data, so to speak, for the rest of the country,” DaCosta said. So when it came to developing a similar tool for more cities, he explained: “It seemed like a very achievable goal.”

Enigma is about four years old. Its business model centers on helping companies, as well as government organizations, improve the ways they organize, explore and use data.

DaCosta described the company’s work as heavily focused on how to get data “into a shape where it can actually start to drive decision-making.” He sees Smoke Signals as just one example of what’s possible when this approach is applied in the public sector. “For us it’s about trying to promote data-driven decision making in government as a best practice,” he said, “and as something that’s very achievable given the state of technology today.”

Bill Lucia is a Reporter for Government Executive’s Route Fifty.

NEXT STORY: Smoking Ban Repeal Vote in Wyoming; Unpaid Water Bills Protest in Detroit Enclave