Using AI to protect people who are at risk from cyclones

Non-profit SEEDS (Sustainable Environment and Ecological Development Society) is using Artificial Intelligence to identify people at risk from cyclones and help move them to safer ground.

This is how the AI model works: It uses high-resolution satellite images to spot the houses in the way of a cyclone; it even charts a risk score for each house as per a dozen parameters.

It was put to effective use during the recent Cyclone Yaas in Odisha, TOI reports, with SEEDS reaching out to a thousand families from communities at risk.

The AI model was created by data scientists at Microsoft and Bengaluru-based Gramener, and is built on Microsoft Azure. SEEDS also got technical and financial support from Microsoft’s AI for Humanitarian Action.

“We usually come to know a cyclone’s predicted path 48 hours or so before it makes landfall. We can run the model before the disaster and issue warnings to people who are at maximum risk,” Raghav Raghunathan, technical director, SEEDS, told TOI.

Mridula Garg, urban researcher and architect on the project, said the risk score results could be obtained within a day, whereas earlier they had to use volunteers for manual survey.

Kate Behncken, lead of Microsoft Philanthropies, said that the solution will help protect people better for ‘other severe weather scenarios’.

And that’s what SEEDS is planning next: trying to create a replica of the model in order to identify people who are at risk during heatwaves.