Drone footage shows the extent of damage done by Hurricane Ian on Sanibel Island. Hurricane shutters on one home were just blown through like it was tin foil.
COLLEGE STATION, Texas — When a major hurricane like Helene or Milton leave wide swaths of destruction, it can be overwhelming and time-consuming to fully grasp the extent of the damage and accurately gauge how long of a recovery process is ahead.
Researchers at Texas A&M are hoping to quicken the process using a combination of artificial intelligence and machine learning.
They have spent more than a year studying damage photos taken via drone from 10 major disasters, including the devastating hurricanes of Harvey, Michael, Ida, Laura, Ian and Idalia.
The team then recruited 130 high school students from Texas and Pennsylvania to label the level of damage for 21,700 buildings on 16,500 acres of land and 400 miles of roads.
The data was used to train an AI system to recognize what storm-damaged infrastructure and roads look like.
With the new system, researchers say if they can get drone video of an affected neighborhood, they can have a damage analysis ready in only four minutes, just by using a laptop.
“AI offers tremendous value for rural counties which do not have the budget or workforce to conduct physical damage assessments but do have inexpensive drones,” said Dr. Robin Murphy, the lead author of the study and Raytheon Professor in the Department of Computer Science and Engineering.
Murphy’s team has already used their system to help the state of Florida in the wake of Hurricanes Debby and Helene, and are on standby to assist with Hurricane Milton damage.