Flood Severity AI: The Use of Iseechange Crowdsourced Data and AI to Manage Public Works, Emergency Response, and Stormwater Mitigation Planning

Flooding has caused over $850 billion in damages in the United States since 2000, accounting for two-thirds of the costs from all natural disasters combined. Despite its significant financial burden and health impacts on frontline communities, especially those disproportionately affected by flood risk and pollution, data on chronic minor flooding incidents remains sparse.

ISeeChange (ISC) is an AI-enabled data and community engagement platform designed to help cities adapt to climate change. By harnessing real-time, hyperlocal data generated by residents, ISC empowers cities, counties, and utilities to better maintain, manage, and protect infrastructure, enabling them to both respond to and prevent extreme climate-induced damage. ISC has successfully gathered, processed, and analyzed real-time observational data at a hyperlocal scale, delivering valuable insights to public sector clients, including the City of Miami, Miami-Dade County, and the City of New Orleans. To date, ISC's data has led to $25 million in green and stormwater infrastructure investments in low-income communities, where traditional stormwater models often underestimate risk.

With support from a Small Business Innovation Research grant from the National Science Foundation, ISC developed and launched Flood Severity AI, a tool designed to help public works departments, emergency managers, and stormwater utilities manage flood reports more efficiently and prioritize responses in real time. In June 2024, Flood Severity AI was deployed for the first time to assist the City of Miami and Miami-Dade County during a storm that brought over 12 inches of rain to the region within 48 hours. The ISC platform provided crucial ground truth and situational awareness of flood conditions throughout the storm, offering actionable data for damage assessment and recovery across 14 municipalities and unincorporated areas of Miami-Dade. ISeeChange was ingested by Miami Dade County’s Emergency Operations Center as well as used by the Department of Public Works and Transportation to deploy vac trucks and temporary pumps. Notably, ISC captured data from up to 35 census tracts that had never been recorded by traditional 311 systems. Flood Severity AI identified flooding in over 15% of reported posts, with 41% of the data originating from communities in the 90th percentile of the EPA’s EJScore Demographic Index. The AI's flood severity scores strongly correlated with traditional tools like Crisis Track’s damage assessment reports, collected days or weeks after the event, suggesting the potential for ISC data to be used in dispatching field teams for on-the-ground verification of local impacts and faster emergency disaster declarations. Finally, the flood event was particularly pronounced in two Adaptation Action Areas of Miami Dade County, where the Army Corps of Engineers, local municipalities, the South Florida Water Management District, and the County’s Office of Resilience plan significant stormwater infrastructure investments. The data will be used by the Office of Resilience in local community engagement activities as well as future project design.