Over the past week, the Kincade Fire in Sonoma County and the Getty Fire in Los Angeles have burned hundreds of square miles, forcing residents to flee their homes. That’s just the latest outbreak in what’s become an annual season of devastating destruction in California, as dry climate and high winds turn errant sparks from things like downed power lines into massive conflagration. At this point, both fires are yet to be contained.
To get ahead of such blazes, fire department officials in the Bay Area’s Contra Costa County intentionally set four test fires of their own this summer. Each was surrounded by a set of field sensors capable of measuring temperature and humidity. When the initial ignition was detected, the sensors relayed the location of the blaze to a remote dashboard created by Zonehaven, a cloud-based analytics application that incorporate current weather conditions to devise a simulation of how quickly the fire would spread over the next five to 10 hours, and, if left unstopped, what nearby areas would be most at risk.
The application is capable of sending out an immediate alert about an ignition situation to firefighters and local governments—anyone who needs instant information to start battling the blaze and plan evacuations. As you can see in the video below, the concept is being developed with the cooperation of the nearby Moraga-Orinda fire district, which battled California’s deadly Rim and Camp fires in recent years.
“There are two kinds of areas [that we identify],” says Zonehaven CEO Charlie Crocker. “One is early warning. So how can you know that a fire coming? Can you know five minutes earlier [or] 10 minutes earlier? Time is really of the essence. Then number two is how do we get people out of harm’s way? The sooner that we know and the more prepared we are before the event happens, the more likely we’ll be to get more people out, and we’ll save lives.”
The most important part of that idea isn’t the sensor tech. Crocker points out that many companies are focusing on new ways to detect wildfires, from using drones to infrared satellite imagery to training AI to spot smoke signatures on satellite maps. “We’re focusing now more on the evacuation planning and the modeling,” he says. “What are the critical infrastructure in that zone? Where are the community refugee areas in that zone? Where that choke points on the map, and where should they be applying their resources?”
In the past, many departments used static maps and generalized plans, obviously not the most effective way to respond to dynamic disaster scenarios. “We’re focused on giving the community a voice in the evacuation planning process and helping the many different agencies that are involved work from a common playbook,” says Crocker. Should the tests show promising results, the plan is to deploy the sensors in the most sensitive areas, under power lines, where many fires start.
In late October, the global data analytics and security company Splunk invested an undisclosed amount in that idea. Splunk backed Zonehaven through its newly formed $50 million social impact fund for companies using data to improve the world. “We see this ecosystem of digital indicators of wildfire developing,” says Tim Woodbury, Splunk’s director of State and Local Government Affairs. “We see tech folks to come up with ideas and then not really having listened to what the real needs of the actual people on the front line issues need. That is not the case with Zonehaven. The idea for the company is coming almost directly from these fires chiefs and is exactly what they need right now.”