ASCE/AWWA/FES Gainesville Branch September 2026 Lunch & Learn

Presentation Description- Increase in frequency and severity of wildfires has resulted in greater harm to human lives, civil infrastructure, and the economy. An essential step toward building wildfire resilience requires deep understanding and robust modeling of human behavior before, during, and after wildfire events. Research on harnessing big data and AI to enhance wildfire evacuation analysis and modeling is emerging but still in its infancy. To address this research gap, this presentation first introduces an analytical framework that systematically extracts individuals’ evacuation behaviors (including whether, when, where, and how to evacuate) from large-scale GPS data collected from mobile devices. Then, two AI models are introduced: One develops situational aware deep learning models to forecast real-time travel demand during wildfires, and the other uses behavioral theory-informed large language models to represent evacuees’ decision-making processes.

Presenter Description- Dr. Xilei Zhao is Associate Professor and Director of AI Initiatives in the Engineering School of Sustainable Infrastructure & Environment (ESSIE) at the University of Florida. Dr. Zhao received a Master’s degree in Applied Mathematics and Statistics and a Ph.D. degree in Civil Engineering from the Johns Hopkins University in 2017. She specializes in developing trustworthy AI and data science methodologies to tackle challenging problems in disaster resilience and transportation systems. Her research has been recognized by multiple awards, e.g., the 2024 NSF CAREER Award, the National Academies’ Early-Career Research Fellowship, and the 2020 Travel Behaviour and Society Outstanding Paper Award.