Research

We develop physics-based and data-driven digital twins of buildings and infrastructure systems using structural sensing, data fusion, and health monitoring. Our research focuses on extracting system-level information from sparse and heterogeneous data to characterize damage, functionality, and performance before and after extreme events.

These digital twins support rapid diagnostics, scenario evaluation, and uncertainty-aware decision-making, particularly in post-disaster environments where data may be incomplete or degraded. Applications include instrumented buildings, transportation infrastructure, and critical facilities.

Infrastructure systems are increasingly exposed to interacting and sequential hazards. Our research addresses compounding hazards—including earthquakes, extreme wind, flooding, fire, and climate-driven stressors—by modeling how damage and disruptions cascade across infrastructure networks and communities.

We quantify risk, reliability, and recovery under multi-hazard scenarios while accounting for interdependencies among structural systems, lifelines, and social functions. This work supports hazard-informed planning, mitigation prioritization, and resilience-based design.

Resilience is not solely a physical problem. We investigate cyber-physical-social systems that integrate infrastructure performance, sensing and data platforms, and human and institutional decision-making processes.

Our research examines how information flows, governance structures, and community behavior interact with physical infrastructure performance, influencing recovery trajectories, equity outcomes, and policy effectiveness. This systems-level perspective bridges engineering analysis with real-world decision contexts.

We translate complex models into actionable insights for engineers, planners, and policymakers. Using community resilience modeling, recovery simulation, and scenario analysis, we support decisions related to mitigation investment, retrofit prioritization, and post-disaster recovery planning.

Our work is applied in real communities through partnerships with federal agencies, state and local governments, and resilience platforms, ensuring that analytical tools directly inform practice rather than remain purely academic.

We develop artificial intelligence and data science methods tailored to natural hazards and infrastructure resilience problems. Our work emphasizes physics-informed and interpretable approaches that integrate machine learning with probabilistic modeling and uncertainty quantification.

These methods support rapid damage assessment, performance prediction, and risk-informed decision-making while maintaining transparency and reliability for engineering and policy applications.

Our research advances structural dynamics and performance-based engineering methods for assessing and designing resilient structures. We investigate system identification, Bayesian filtering, and dynamic response reconstruction to evaluate damage, functionality, and recovery.

This work provides the mechanical foundation for digital twins, post-event assessment, and resilience-based design of buildings and infrastructure systems.