When a disaster strikes a determination must be made to measure if the resources needed to recover are expected to exceed local and state capability. Preliminary Damage Assessments (surveying damages) are conducted to enable States, local jurisdictions and FEMA to determine the magnitude of damage and impact of a disaster event. The process typically takes a minimum of 40 days and on large disasters can take many many months to complete.
Our task was to developed a Preliminary Damage Assessment (PDA) and Decision Support tool to expedite this process to a few hours or days post impact.
Utilizing over 40 databases, machine learning, feature engineering, geospatial engineering and analytical techniques, iParametrics developed a custom probability model that runs within R (Environment for Statistical Computing) that classifies if a county will be declared eligible for Federal disaster funding under FEMA’s PA program. This allows end-users to make rapid, complete assessments for Public Assistance funding during the Preliminary Damage Assessment (PDA) phase of a disaster within hours of the event.
Immediately after a hurricane event (within 1 day), the System identifies the counties to be declared federal disaster areas with a greater than 95% accuracy. This allows planners to begin identifying and mobilizing the appropriate personnel and physical resources in advance of the actual disaster declaration with confidence that the declaration will in fact be made. Further development will allow for predictive analysis based on National Hurricane Center forecast information and additional data sets. It also will generate a cost estimate, by category of work, for the impacted area on a per capita basis.