How artificial intelligence could be used to fight wildfires

It is no secret wildfires are growing more frequent, more expensive to fight and rebuild from, more widespread, and more sustained than ever before. 

A Virginia-based technology integration company called Science Applications International Corp., or SAIC, wants to use a new type of artificial intelligence to fight those fires by using predictive analysis before the fires even break out and then AI logistics to help contain them. 

"They need to understand where fires can occur, where they're occurring and where they're going," SAIC chief climate scientist Stephen Ambrose said. "And even after the fire — what happens after that?"

The new software uses familiar techniques to produce a next-generation outcome by processing and analyzing huge amounts of data, along with human observations, to put answers in the hands of decision-makers in real time, including where to deploy resources and what areas might be in danger. 

"From remote sensing, from satellites, we get it from other types of data — surface data, observations of any type that can go in there and be processed through this technique," Ambrose said.

However, Ambrose said it still isn't enough. SAIC is developing a fourth-generation of artificial intelligence the company calls artificial intuition. An algorithm actually replicates the human brain's thought process. That means the program could eventually learn to think for itself, saving time, money, and lives. 

"We're looking more at the way the human brain thinks," Ambrose said. "This technique that we have can do that — so it takes a lot less data. It looks at the complexity of that data."

In 2021, authorities spent nearly $4 billion fighting wildfires in the United States, nearly double the amount spent the year before.