By Ines Aviles-Spadoni, M.S., UFTI

Dr. Xilei Zhao is an assistant professor in UF’s Department of Civil and Coastal Engineering.  She has been awarded an NSF Faculty Early Career Development (CAREER) grant titled “An Integrated Trustworthy AI Research and Education Framework for Modeling Human Behavior in Climate Disasters.”  This grant will support research in developing and using AI responsibly based on a White House Executive Order.

“This CAREER work will develop and apply trustworthy AI methodologies to fundamentally transform behavioral modeling in climate disasters and advance the state of knowledge about the complex dynamic protective action decision-making,” Dr. Zhao said.

This innovative project is crucial for saving lives in the face of our current climate crisis, where severe weather, storms, wildfires, and hurricanes have become increasingly common. These natural calamities leave a trail of disaster in their paths, literally uprooting lives, damaging infrastructure, and impacting economies.

However, there is hope on the horizon with AI modeling techniques that help researchers understand and forecast human behavior in climate disasters at a granular level. What does this mean? Researchers want to know how all types of people behave in a variety of situations, regardless of where they live, and when their fight-or-flight instincts kick in, such as when they are trying to flee from a wildfire. This research could lead to more effective and timely evacuation strategies to make sure that people from all walks of life are safely out of harm’s way.

What sets this project apart is its unique approach. Unlike current AI models that overlook essential principles such as trustworthiness, fairness, or bias in their algorithms, Dr. Zhao and her research team are committed to developing a more comprehensive and responsible AI model. This approach aims to avoid the unintended consequences that may arise from using off-the-shelf AI algorithms in disaster management, potentially saving more lives in the process.

“The CAREER research will create theory-informed, high-accuracy AI-based behavioral models that are explainable and with bias minimized to inform high-stakes decision-making in emergency management,” Dr. Zhao said.

Just think about the lifesaving benefits this new technology can offer. It could potentially be applied to mobile apps used for wildfire or hurricane evacuation and to computer programs used by emergency management center officials. The ultimate goal is for people and emergency management officials to use this new technology for efficient and safe evacuations.

Dr. Zhao’s research will also have an educational and workforce development component. The results will be added to STEM curricula available to Florida high school students and existing engineering curricula for undergraduate and graduate students. They will also benefit emergency managers and transportation professionals by training them to increase their skill sets.

This project is an example of what AI can do for the greater good, ensuring that lives are saved in the face of these ever-increasing natural disasters.