Coupled human-infrastructure systems capture the intricate interdependencies between communities and the infrastructure they rely on. These systems are shaped by how individuals and communities use infrastructure, depend on it, and experience its failures, especially during times of disruption. Disasters often expose disparities in resilience, as households—such as those with limited mobility, economic resources, or access to services—face heightened risks due to their constrained ability to prepare for, respond to, and recover from events.

My research focuses on developing modeling frameworks that simulate these interdependencies to better understand the dynamics of socio-technical systems. Using system dynamics and complex network-based approaches, we explore both infrastructure-to-human and human-to-infrastructure feedback during disasters. This integration enables us to analyze how human behavior affects infrastructure performance and vice versa, ultimately offering insights into how targeted interventions can enhance community resilience.

Households depend on infrastructure systems for essential services, and during disruptions, they often adapt by adjusting their demand or supplementing supply to lessen the impact. These adaptations—whether planned or spontaneous—play a critical role in maintaining household functioning and resilience, serving as a bridge between infrastructure performance and community well-being during crises. Understanding these behaviors is vital for supporting equitable and effective disaster response.

This research aims to characterize and predict household adaptations during infrastructure service interruptions using econometric models and machine learning techniques. By capturing how and when households implement various adaptations, the study provides insights for system operators, emergency managers, and community organizations to better incorporate adaptive behaviors into planning, restoration efforts, and resource allocation.

Critical facility accessibility is essential for disaster response and recovery, yet road networks often provide uneven access across communities. Disruptions such as flooding can isolate populations from essential services or significantly increase travel times, disproportionately affecting communities at risk. This research adopts an integrated network-based approach to evaluate accessibility by accounting for partial road failures, multiple types of critical facilities, and changes in travel time during flood events.

Some roads play a disproportionately important role in maintaining access to these critical facilities. To increase disaster resilience, my work involves identifying the most critical roads for protection and early recovery. While certain segments are frequently used in post-flood travel routes, others are vital for preserving access to avoid formation of service deserts. My research combines complex network-based approaches with optimization techniques on large-scale networks to identify and prioritize essential roads for protection and early recovery.