Assessing Disaster Impact in Real Time: Data-Driven System Integrating Humans, Hazards, and the Built Environment

Abstract

Current disaster impact assessments are conducted within days after an event and usually with limited workforce and resources. Although many recognize the valuable information attainable in this way, the evidence is likely to be disturbed, and collecting information rapidly is critical for timely disaster response. Many proposed approaches use either physical- or social-sensing data, and are not implemented in real time, and therefore cannot assess disaster impacts holistically and timely. It also is challenging to integrate multisource data with disparate data modalities and resolutions, and to accommodate large-volume, high-rate input data for real-time computations. Therefore, we propose an assessing disaster impact in real-time (ADIR) system which harnesses cutting-edge computation and geovisualization platforms. ADIR processes heterogeneous data for humans, hazards, and built environments, and outputs multiscale disaster impacts. The usefulness and effectiveness of ADIR were demonstrated in the city of Houston, which was affected by Hurricane Harvey. ADIR provides a platform for multiagent coordination during community response and recovery, which contributes to more-resilient communities by potentially reducing life and property losses.