A University of Texas at Arlington architecture professor is using artificial intelligence to make sure people living in disaster areas have a safe way out.

"Integrating infrastructure damage modeling with transportation systems means we're not just routing people, we're routing them through paths that are actually usable, safe, and accessible in real time," Assistant Professor Mahmoud Bayat said in a story by The Shorthorn, the student newspaper.

At issue is the growing number of extreme weather events happening around the world. NASA has linked our planet's overheating to increased risks for life-threatening floods, wildfires, and other disasters. National Oceanic and Atmospheric Administration data is in line with the NASA assessment. There have been 115 billion-dollar disasters — cyclones, tornadoes, wildfires, and droughts among them — during the last five years (2020-2024), resulting in more than 2,500 deaths. There were 33 such events in the 1980s, per the report. The valuation was adjusted for inflation.

The severe storms can wipe out key infrastructure, including escape routes. Bayat's AI-powered model uses updated data to account for risks, transportation access, and other factors, per The Shorthorn.

Calculating bridge stability in Galveston, Texas, is an upcoming part of the pilot project. Disasters will be simulated using live traffic, structure, and weather information.

"You have to consider social vulnerability of flooding and hurricanes — and also the equity of the community to access the transportation network," Bayat said in the report, noting that ensuring elderly people, and folks without cars, are not forgotten during evacuations.

AI doesn't come without a cost. Goldman Sachs reported that data center demand is expected to increase by 165% by 2030, driving huge increases in electricity draw. The Massachusetts Institute of Technology added that the powerful servers require loads of water for cooling.

The report cited research by Google and the University of California at Berkeley that estimated enough power to energize 120 U.S. homes for a year is needed to train and deploy a service such as OpenAI's GPT-3, producing about 552 tons of planet-warming carbon dioxide. Those are the types of fumes contributing to severe weather risks that Bayat's program is troubleshooting.

The overheating potential can be limited with investments in renewable power. Meta and other companies have announced cleaner solar projects. Microsoft is even planning to reopen a portion of the infamous Three Mile Island nuclear facility in Pennsylvania to help power its tech operations. That project has drawn criticism, as well.

At UT Arlington, Bayat's research is mainly focused on bridge safety and traffic loads. He cited aging Texas spans in the story. It's part of the way the expert envisions leveraging groundbreaking technology for the greater good, according to The Shorthorn.

Do you think your city has good air quality? Definitely 🥰 Somewhat 😮‍💨 Depends on the time of year 😷 Not at all 🤢 Click your choice to see results and speak your mind.

Other innovations are using AI and even augmented reality to predict when and how storms will hit communities.

"Equity in evacuation planning ensures that no one is left behind. Our system intentionally accounts for these vulnerabilities, modeling evacuation options," Bayat said in the story.

Talking with friends and family about AI — and its impact on society and weather prediction — can spotlight the importance of how tech can both help and hinder society. A broader understanding of climate challenges can highlight ways to help, including supporting planet-friendly brands. It's also important to hold companies such as Meta, Microsoft, and others accountable for their sustainability goals, ensuring they aren't talking a big game without following through on the pledges as they continue to develop power-hungry data centers.

Join our free newsletter for weekly updates on the latest innovations improving our lives and shaping our future, and don't miss this cool list of easy ways to help yourself while helping the planet.