How insurtech is powering wildfire response

Register now

As wildfires continue to scorch sections of the country, start-up insurtechs and existing tech companies have been working to develop tech innovations addressing extreme-weather disasters. Investors have stepped in to fund emerging efforts around cleantech, a term used broadly to describe technology seeking to manage human impact on the environment. While hopeful, the technology can take years to prove and even longer to convince traditional utilities and government agencies to adopt.

Satellite Imagery
Some start-ups are ambitiously looking to tackle wildfires head-on. Chooch AI, an artificial intelligence company based in San Francisco, is using a system that analyzes satellite images every 10 minutes to identify where new wildfires may have broken out. At the moment, firefighters largely rely on traditional methods to spot fires – typically people who call in after seeing smoke. That can lead to false alarms and fires that go unnoticed for too long, said Chooch CEO, Emrah Gultekin. Chooch’s technology is trained to spot likely signs of wildfires, and then send photo emails to people at firefighting agencies, who can then verify if a fire has broken out. Gultekin says they’re talking with the California Department of Forestry and Fire Prevention (Cal Fire), as it adapts its system to look for wildfires, although it’s early in the process. The company hopes to have the system live by next year’s wildfire season.

The Power of Drones
Drone technology has been useful to firefighters for years. But as the technology develops, drones are getting smaller and more powerful. They also have better payload options and can go where humans cannot. An aerial view provides a quick assessment of the situation and allows fire crews to see through smoke and identify hotspots.

Here are some ways firefighting drones help:

  • Scene Monitoring
    • Rapid 360° assessment of burning structures
    • Floodlight illumination of night-time operations
  • Wildland Firefighting
    • Visualize current fire conditions and respond immediately to change
    • Reduce ground crew risk with enhanced monitoring
    • Live stream reliable data to command centers
  • Search & Rescue
    • Survey large areas of inaccessible terrain
    • Payload drop systems deliver items to victims and crew
    • Locate individuals at night with thermal imaging cameras and floodlights
  • Post-Fire or Disaster Assessment Survey
    • Map the scene to assess damage
    • Search for and locate missing people after a natural disaster
    • Use footage for future training purposes

How AI Helps
Descartes Labs, a spin-off of Los Alamos National Lab, uses artificial intelligence (AI) to detect wildfires. Their wildfire detector, launched in 2019, can correctly spot a wildfire in nine minutes. The wildfire detector uses images from two Geostationary Operational Environmental Satellites, GOES-16 and GOES-17, launched by the National Oceanic and Atmospheric Administration (NOAA). These satellites provide a stream of high-resolution and infrared imagery every five minutes. A variety of AI algorithms ensures the tool filters out false alarms, such as flares from oil and gas industries, controlled agricultural burning, or large fireworks displays. The algorithms take into account where refineries are, locations of steel and copper mills, the overall terrain, and more.

For example, when the tool spots a wildfire in California, Descartes Labs immediately broadcasts an alert to the LA Times who, in turn, reports it on their wildfire map. The detector can provide the exact latitude and longitude of the fire. In October 2019, Descartes detected the Kincade Fire in Sonoma County, CA, prompting the LA Times to release the first public notice about it.10

Preventing Wildfires Sparked by Powerlines
California utilities are experimenting with a new technology that proponents say could help prevent both electricity shutoffs and equipment failure related wildfires. The technology comes as the Pacific Gas & Electric (PG&E) faced intense criticism over its Public Safety Power Shutoff program, which led to hundreds of thousands of Californians being left without power, often for days at a time, in an effort to prevent failing equipment from starting wildfires.

Distribution fault anticipation (DFA) is a technology using predictive algorithms to assess electric systems and identify potential equipment failures. B. Don Russell, a Texas A&M University professor, helped develop the technology as part of the Power System Automation Laboratory. “It will be a paradigm shift for the way utilities operate,” said Russell. Currently, electric companies “just kind of have to wait until something breaks and then go fix it,” Russell said. “DFA gives you real-time situational awareness of the circuit. It allows you to know when things start to degrade.”

The original purpose of the technology was to prevent costly power outages and prevent accidental electrocution from downed powerlines. “It just turns out that the things that cause outages in the system are also things that start wildfires,” Russell said. PG&E began testing the technology in 2019, as part of a larger project. PG&E spokesman Paul Doherty stated in a November 2019 Sacramento Bee article that “The technology is being evaluated along with other sensor technologies as a way to detect emerging conditions on the electric grid and improve situational awareness... The evaluation phase is scheduled to be completed by July 2020.”

To learn more about wildfire risk and preparation, Allianz has prepared a report: Future fires - Weathering the fire storm.

For reprint and licensing requests for this article, click here.
Drones Artificial intelligence