Data-AI-Resilience LabNational Central UniversityGoogle Scholar ↗RESEARCH PROGRAM
We combine mechanics, artificial intelligence, and high-performance computing to study infrastructure performance from individual components to entire cities.
WHY IT MATTERS
High-fidelity nonlinear analysis is essential—but applying it to tens of thousands of buildings is computationally demanding.
Our work develops scientifically grounded AI and automation strategies that retain important physical behavior while expanding the scale, speed, and accessibility of structural and regional risk analysis.
CORE THEMES
Machine-learning methods for structural response prediction, damage classification, building inventory generation, and rapid post-event assessment.
Physics-based ground motions, fragility analysis, and regional building data to quantify risk across buildings, communities, and cities.
Frameworks for understanding how infrastructure systems withstand, adapt to, and recover from hazards—and how analysis can inform action.
RESEARCH WORKFLOW
Street-view imagery, 3D point clouds, sensor records, structural drawings, ground motions, and regional building data.
Automated inventories, nonlinear structural models, physics-based simulations, and probabilistic fragility representations.
Deep learning, active learning, generative models, graph neural networks, and transfer learning constrained by engineering knowledge.
Building response, collapse risk, regional loss, recovery, and infrastructure resilience at scales useful for decision-making.
COLLABORATION
We welcome collaboration in structural engineering, AI, statistics, earth science, computing, and urban resilience.
Start a conversation →