Initiatives under way to use AI for better natural disaster detection, response

Private industry and the United Nations have projects under way using artificial intelligence to help communities warn citizens of natural disasters and respond better to them.

A new generation of earthquake monitoring and climate forecasting systems powered by AI is allowing scientists to detect earthquakes in real-time and more accurately predict the path of hurricanes.

Start-up SeismicAI is working in Mexico to deploy an AI-enhanced network of sensors that can detect earthquakes in real-time. Machine learning techniques have been adopted into weather forecasting models for many years, but recent advances give new models the ability to make more expansive use of AI, improving the accuracy and speed of forecasting.

AI is also being used to assess damage after disasters faster and more accurately. For example, following both Hurricane Milton and Hurricane Ian, non-profit GiveDirectly used Google’s machine learning models to analyze satellite images to identify the worst affected areas, and prioritize cash grants accordingly.

AI-driven models are limited by the quality of data they are trained on, though. Poorer regions tend to have fewer and worse-maintained weather sensors, for example, making the models less effective.