Predicting weather extremes with smart data and AI

A US start-up is working on precise weather forecasts using satellite data and AI. How does it work?

Transcript

00:00Tomorrow.io also uses AI. Smart data analysis is supposed to deliver faster and more precise weather forecasts.

00:11A big model that is doing a forecast using a lot of resources takes a few hours because you're discretizing the entire globe.

00:18These AI models can do that in a matter of minutes.

00:22The AI models improve when they receive more data, provided by both company-owned and other satellites.

00:29As we collect more and more of this data, especially the precipitation data, which is really the hardest one to get globally,

00:36we feel that this will be a game-changer in training all of the AI models.

00:41We want to be able to be much better in predicting precipitation,

00:45and that's where I believe that with all of our models and the data, we're going to make a big positive impact.

00:52Tomorrow.io supplies airlines, electricity providers and TV broadcasters with weather forecasts,

00:59which allows them to quickly adapt to severe weather conditions. But others could profit too.

01:08In Kenya right now, we've helped five million farmers understand when to plant, when to harvest to improve their yields.

01:15In the face of drought and more volatile weather, we can deliver forecasts to farmers in the way they need to receive it.

01:21That might be through SMS text or through smartphones or other methods as well.

01:25But they understand based on where they are and based on where their farm is exactly what to do and when to maximize their yield.

01:33Precise weather forecasts help people and companies to prepare for extreme weather and to minimize the damage.