New AI tool can predict which planetary systems can survive

Published by:
Posted on: July 14, 2020 5:35 pm EDT

Artificial intelligence is confidently going where no one has gone before

Artificial intelligence (AI) can be developed in virtually any area of our lives, even to help astronomers with their predictions of the space. One of the biggest challenges for astronomers is the difficulty of calculating if a planetary system is stable. Until now, scientists needed to calculate the motions of multiple interacting planets over billions of years, which takes quite a long time to do. However, a new AI model allows machine learning to combine and analyze simplified models of the planet’s dynamic interactions to accelerate the process.

Even though the same struggle led to the discovery of many mathematical revelations, no one has ever found an effective way to theoretically predict stable configurations. “Separating the stable from the unstable configurations turns out to be a fascinating and brutally hard problem,” said Daniel Tamayo, a NASA Hubble Fellowship Program Sagan Fellow in astrophysical sciences at Princeton. With this new tool, the huge swaths of unstable orbital configurations are taken out of the picture quickly – in a matter of minutes. Something that would’ve taken thousands of hours for a human to do.

According to one of the co-authors of this technology, Phil Armitage, it “demonstrates that by combining our hard-won understanding of planetary dynamics with modern machine learning techniques, we can reliably predict the fate of an abundant class of known extrasolar planetary systems.” It basically takes away any calculation that is known to be unstable and save time having to go through the only ones that have a chance of being stable. This model can simulate around 10,000 orbits in only fractions of a second.