AI model predicts climate disaster risks to homes and emergency routes
A new AI can forecast which buildings are most vulnerable to floods and heatwaves.
Deep Dive
Researchers developed an AI model called Skjold-DiT to predict climate risks like flooding for individual buildings. It analyzes urban data, including transportation networks, to forecast hazards and their impact on emergency vehicle access and evacuation routes. The system was trained on a dataset of over 847,000 building observations across six cities. It provides uncertainty-aware forecasts to help cities plan for disasters and improve emergency response logistics.
Why It Matters
This technology helps cities protect residents and plan smarter emergency responses to climate disasters.