Current disaster damage assessments are conducted days following an event and usually with limited workforce and resources. Valuable information is gathered through disaster assessment, but this evidence is perishable, and likely lost once the population returns and begins repairs. A critical need exists to collect information rapidly to preserve it for later analysis.
The FIBER research team is using emerging data types (e.g. textual and visual data) and sources (e.g. social media and crowdsourcing disaster mobile apps) and machine learning algorithms to speed up damage estimation at multiple scales. The team includes collaborators and partners from the City of Gainesville-Smart City Office, local emergency management, and partners inside and outside UF.