Wildfires are increasing in frequency and intensity as climate change becomes more pronounced and visible. These are now causing disruptions in urban areas, as seen in __, where 1000s of people left homes and their lives behind. Now, they'll have to wait anxiously to know the state of their homes and the damage that they'll need to fix. Sometimes, this wait prolongs to weeks and months.

Experts at Stanford University and the California Polytechnic State University (Cal Poly) have developed an artificial intelligence (AI) algorithm system called DamageMap that is a damage classifier; it helps identify building damages within minutes of a catastrophe by studying aerial photographs.

“We wanted to automate the process and make it much faster for first responders or even for citizens that might want to know what happened to their house after a wildfire,” said lead study author Marios Galanis, a graduate student in the Civil and Environmental Engineering Department at Stanford’s School of Engineering in an official press release by Stanford. “Our model results are on par with human accuracy.”

Previous systems would need pre-and-post damage photos and run a comparative test between the two. This has restricted rapid assessments since the system requires data from the same satellite, camera angle, lighting conditions, etc which many a-times would be either impossible or time-consuming. However, unlike previous systems that did comparative analyses between pre-and-post-fire photographs, DamageMap is a trained program that uses machine learning to study post-fire images so, it can adapt to sifting through unseen data.

“People can tell if a building is damaged or not – we don’t need the before picture – so we tested that hypothesis with machine learning,” said co-author Krishna Rao, a graduate student in Earth system science at Stanford’s School of Earth, Energy & Environmental Sciences (Stanford Earth). DamageMap studies structures for damage indications such as blackened surfaces, crumbled roofs or the absence of structures to estimate damages. “This can be a powerful tool for rapidly assessing damage and planning disaster recovery efforts.” The researchers tested it using a variety of satellite, aerial and drone photography with at least 92% accuracy.

The current approach to understand damage assessment involves a detailed door-to-door survey. While DamageMap may not replace this method of damage classification, it could be used as a scalable supplementary tool by offering immediate results and providing the exact locations of the buildings identified. While DamageMap can not yet participate in detailed surveys to report the extent of damage, it can help teams prioritize areas to survey based on simply determining if fire damage was present or absent.

“With this application, you could probably scan the whole town of Paradise in a few hours,” said senior author G. Andrew Fricker, an assistant professor at Cal Poly, referencing the Northern California town destroyed by the 2018 Camp Fire. “I hope this can bring more information to the decision-making process for firefighters and emergency responders, and also assist fire victims by getting information to help them file insurance claims and get their lives back on track.”

As with all machine learning systems, DamageMap will only improve with time as it consumes data. In the future, the researchers hope that the open-source platform can be applied to assess damages done by other disasters, such as floods or hurricanes. “So far our results suggest that this can be generalized, and if people are interested in using it in real cases, then we can keep improving it,” Galanis said.

The findings appeared in the International Journal of Disaster Risk Reduction. Galanis and Rao developed the project during Stanford’s 2020 Big Earth Hackathon: Wildland Fire Challenge. They later collaborated with Cal Poly researchers to refine the platform, a connection that resulted from Rao and Frickers’ participation in Google’s 2019 ″Geo For Good” conference, where the two built an initial prototype as part of the conference Build-A-Thon