Data Blanket drones are deployed over a controlled fire. Using AI, the devices don’t require a pilot for navigation or to manage battery life. (Data Blanket Photo)

Data Blanket is a new startup out of the Seattle area developing small fleets of AI-powered drones that can perform self-guided missions to assess wildfires and automatically share that information with incident commanders and pilots battling forest blazes.

The Bellevue, Wash.-based company launched in March 2022 and has been providing demos for firefighting agencies, showcasing the drones’ ability to share real-time images of a fire from a bird’s-eye view.

Or as one fire chief told CEO Omer Bar-Yohay, “You just made my firefighters 400-feet tall.”

Fires are raging worldwide this summer. The ongoing wildfires in Canada have for weeks pumped smoke into the U.S. Fires in Greece and Italy scorched Mediterranean Islands and forced evacuations. The Pacific Northwest has seen hundreds of mostly small blazes, but is heading into peak wildfire season.

The hope is that technology like Data Blanket’s will help firefighters safely and quickly size up a wildfire. “It’s a game-changer in the way we manage and identify what we’re up against,” Bar-Yohay said.

Founders and funding

Data Blanket was co-founded by Bar-Yohay and Chief Operating Officer Yair Katz, both former executives at Eviation Aircraft, an electric aviation company that Bar-Yohay also co-founded and led. Years ago, the two served together in Israel’s military, where Katz led special operations for the army.

The startup’s third co-founder is Gur Kimchi, who co-founded and for seven years led the Amazon Prime Air delivery-by-drone project. Among other roles, Kimchi worked for a decade at Microsoft on projects including Virtual Earth and search. He is a founding member of the Federal Aviation Administration’s Drone Advisory Committee and serves as a Data Blanket board member.

Image from a drone camera detecting a fire using an AI-based computer vision system. (Data Blanket Photo)

The trio bring together wide-ranging expertise that encompass skills including creating startups, running tactical operations, and an understanding of the FAA’s drone regulations.

Their backgrounds and efforts to date have already attracted the attention of notable backers. Earlier this year, the company raised more than $4 million in seed funding from investors including Bill Gates’ Breakthrough Energy Ventures and Innovation Endeavors, which was co-founded by former Alphabet CEO Eric Schmidt, Dror Berman and others.

“Losses from catastrophic wildfires are increasing at unprecedented rates, and at the same time, our emergency responders are left fighting these fires in the same way they have been for the last 50 years,” Berman said via email. “Data Blanket is bringing firefighting, and ultimately wildfire management, into the digital age.”

The tech

Data Blanket’s strategy is to deploy drones outfitted with RGB and infrared cameras, AI-based computational software, 5G and Wi-Fi communication capabilities, and advanced navigational features.

The drones are programmed to navigate and inspect a fire area, launching and landing on their own, without guidance from pilots. One onsite operator per eight UAVs is needed to ensure the vehicles operate safely and don’t interfere with manned aircraft, which the system can detect. A swarm of four drones can survey about a 2-mile radius.

The drones can establish 5G and Wi-Fi networks to communicate with local devices and off-site users implementing a satellite communication link. The UAVs gather real-time 3D information about an area’s topography, vegetation and fuels, which is fed into the startup’s software or other firefighting management platforms. The data are used to map the perimeter of a fire and can identify areas that have burned or still burning. The system can also overlay locations of firefighting crews and equipment.

Real-time data from the drones are fed into Data Blanket’s software to map the perimeter of a fire. Other relevant fire incident information is automatically merged into the app and displayed as 2D or 3D images. (Data Blanket Image)

Fires are fought by identifying their boundaries. A wildland blaze isn’t extinguished by dousing it with water — unless the rain complies. Rather it’s contained by creating boundaries that won’t burn and the fire eventually consumes the available fuel within that area.

In the near future, the Data Blanket system will be able to generate micro-weather information, which can be dynamic in the area of a fire. The team plans to use data gathered from fires to better predict how a blaze will behave.

Data Blanket is using drones from Ascent AeroSystems, a Massachusetts-based manufacturer of high performance, unmanned aerial vehicles (UAVs). Kimchi joined the company’s board in 2020.

UAVs in firefighting

While Data Blanket aims to distinguish itself by its use of AI, drone-assisted firefighting is well underway.

At first it was wildfire hotshot crews buying toy drones and modifying them, said George Geissler, Washington’s state forester and the Department of Natural Resources’ deputy supervisor for wildfire operations.

Now there are numerous drone companies and startups selling devices for fighting and recovering from wildfires. Seattle-based Mast Reforestation (formerly known as DroneSeed) operates drones that disperse tree seeds across charred landscapes. The state has its own industrial drones that it deploys to assess wildfires and perform prescribed burns.

“Not only are you getting better visibility on fires, you don’t have to send a human being into an area that is potentially hazardous,” said Geissler, speaking in general terms about drone use.

But he has some concerns about all of the technology that’s flooding the space. It’s essential, Geissler said, that the tech can smoothly integrate into existing management systems. He’s concerned about overwhelming commanders and firefighters with too much information, making their jobs harder instead of easier.

A swarm of Data Blanket drones. (Data Blanket Photo)

‘Around the clock’

Bar-Yohay acknowledges the competition in the field and reiterated that Data Blanket’s system doesn’t require trained pilots to fly the UAVs.

“It’s a different way to fly and operate swarms of drones, teaching a machine to manage its resources based on what it needs to see,” he said.

Data Blanket is working with the FAA on two waivers: one will allow anyone to operate their system without restrictions, and another allowing the drones to fly beyond where an operator can see it — an exception known as “beyond visual line of site” or BVLOS.

Katz expects approval for both waivers “very soon.” There’s a case, he said, for changing FAA regulations for advanced systems like theirs used to fight wildfires, but he doesn’t support a widespread weakening of rules, which could be dangerous.

The startup, which has 14 full-time employees, is currently demoing its devices with fire agencies on the West Coast and fine-tuning the system features. The plan is to start releasing the products next year. BaYohay said customers, namely government-run fire departments, could buy Data Blanket’s system or use it as a service by the hour or day.

The interest, he said, keeps growing.

“The crisis in wildfires and forest management is so intense that it’s around the clock. We get requests and comments and people needing more literally every day,” Bar-Yohay said. “I really hope we can deliver in a way that makes a difference.”

Like what you're reading? Subscribe to GeekWire's free newsletters to catch every headline

Job Listings on GeekWork

Find more jobs on GeekWork. Employers, post a job here.