Problem Statement

During floods, warnings may indicate that an area is at risk, but conditions can change rapidly from one street to another. Roads may become flooded or blocked, while other routes remain accessible. People and emergency responders therefore need timely, location-specific information about actual road conditions, not just a general flood warning. There is a need for a system that predicts flood risk, detects changing ground conditions, and alerts the public and authorities with actionable information.

Solution

We will build dynamic flood safety system using a Predict → Detect → Alert workflow. It will predict flood risk, detect changing road conditions using available observations, and send location-specific alerts to the public and emergency authorities,including dangerous and currently accessible system

In Detail

Our system follows a Predict → Detect → Alert workflow. First, it uses rainfall forecasts, historical flood data, elevation, water levels where available, and geographic information to estimate flood risk at area and street level. During the event, available observations such as cameras, sensors, rainfall measurements, and verified reports help detect actual road conditions. Roads can be classified as safe, moderate-risk, dangerous, or unverified. When critical conditions are detected, the system generates location-specific alerts for the public and emergency authorities and identifies currently accessible routes for evacuation and emergency response.