NASA has built an early warning system that can detect solar storms 30 minutes in advance so that preparations can be made. Like most AI systems, it is a machine learning model. Solar storms can damage sensitive electronic devices on Earth. However, they aren’t easy to predict.
Machine learning models are trained with data that teaches them how to detect things automatically, which can include objects or more useful to this case: patterns and anomalies. There are many variables to monitor, many calculations to make, and it is difficult for a human to keep up with all of that without missing details.
The 30-minute warning provides just about enough time to prepare for impacts on electricity grids and other important infrastructure. Solar storms are coronal mass ejections — the expulsion of plasma and magnetic fields from the sun’s corona.