Assessing Roadway Flood Risk with Crowdsourcing and AI

Floods can disrupt urban road networks by slowing down traffic, forcing reroutes, changing daily travel routines, and loss of life. This study uses crowd-sourced data, INRIX traffic data, and hydrologic information and AI models from Houston, Texas, to identify significant risk of flooding. The goal is to uncover hidden travel patterns, assess rerouting behavior, and provide data-driven insights to support intersection-scale roadway flood warnings and management. The framework supports improved understanding of driver behavior during floods, real-time tracking of flood impacts, prioritizing high-risk areas, and planning targeted interventions to improve resilience.