New AI Tool Speeds Up the Hunt for Hidden Planets

By Space.Fan Editorial DeskSeptember 18, 2026 at 11:01 AM UTC

Scientists created an AI that finds baby planets hiding in space dust, making it easier to study how new solar systems form.

Image · Space.Fan

The Drop

Scientists have built a smart computer program called DPNNet-3.0 to help find young s. When s are born, they are surrounded by a flat, spinning ring of gas and dust called a protoary disk. Astronomers often see dark gaps in these rings, which suggest that a baby is orbiting the and clearing a path through the dust. However, figuring out the exact size and mass of these hidden s is a very slow process that requires complex math for every single image. To speed this up, researchers trained this new computer tool using thousands of simulated images. By looking at these pictures, the AI learned to recognize the patterns created by gravity from one or more s. This allows the tool to guess if a is there and how much it weighs just by looking at a picture of a dust disk. To make sure it works, the team tested the AI on real pictures of a system called HD163296 taken by the ALMA telescope. The computer's guesses matched up well with what experts had already figured out using older, slower methods. This proves that the AI can handle complex systems where multiple s might be playing tug-of-war with the dust.

Why It Matters

This tool is a huge time-saver for space researchers. Previously, calculating the mass of a in a dust disk took a massive amount of computing power and many hours of work. Now, the AI can look at a new image and make a strong estimate almost instantly. This helps astronomers survey many more s quickly to see how common s are in our galaxy. By knowing the mass of these young s, scientists can better understand how solar systems grow up. It helps them piece together the story of how rocky worlds like Earth might form out of swirling clouds of space rubble.

The Catch

While the AI is fast, it is only as good as the pictures it was trained on. Because it relies on computer simulations to learn, it might struggle if it sees a system that looks very different from what it practiced on. It provides a smart estimate, but astronomers still need to verify its findings with more detailed physics calculations to be absolutely sure about a 's weight.

Put That in Perspective

In the past, astronomers had to build custom models for every single planet candidate they found, which limited how many stars they could study. This new approach moves us toward a future where computers can automatically scan thousands of telescope images to find new worlds, helping us move from looking at individual examples to seeing the 'big picture' of how planets form.

Source September 18, 2026

ApJ