The solar system has eight planets moving around it, and in a long time, none of them can collide with each other, which is perfect for everyone on Earth. But do other solar systems have planets which might collide with each other?--this is where artificial intelligence would come in.
Analyzing billions of exoplanets can't be done alone
Princeton researchers have developed a new AI that can analyze high-speed data to determine whether the planet's orbits are stable, and which ones will end in catastrophe. It will be a great benefit to astronomers because it will benefit them pin down the planet's orbits in solar systems that they can not observe in adequate detail.
The technology we currently have for exoplanet detection right now is not enough to offer orbital details in perfect precision. Still, if we evaluate what we already know and go through the different choices, we could get a general mechanics idea.
It's possible to study many orbits, however, and going through billions of them would take a long time, even with the most efficient supercomputers.
The Artificial Intelligence That Examines Exoplanet Orbits
The question described above is why Daniel Tamayo, a NASA Hubble Fellowship Program Sagan Fellow in astrophysical sciences, created an algorithm that can be an alternative to the brute force computing commonly used by researchers.
It takes a lot of the energy of a computer to distinguish theoretically stable orbits from unstable ones, even with how powerful modern supercomputers are. There are billions of orbits to move and analyze.
Tamayo's Stability of Planetary Orbital Configurations Klassifier (SPOCK) system, simplifies the process by integrating machine learning techniques with a simplified model of planetary interactions.
The method lets SPOCK figure out which orbits are the most unstable, and it gives thousands of possible orbits in a few milliseconds instead of hours.
The SPOCK algorithm divides the structures into those that will break apart or smash out of those who will stay stable shortly. In this scenario, the Near future means just a few million years. Considering how long a solar system can last, astronomers are unlikely to encounter any of these chaotic orbits' effects.
The artificial intelligence begins its cycle by simulating ten thousand orbits, generating ten summaries of data from the simulations, and capturing the system's resonant dynamics.
The algorithm then produces predictions based on the data on whether after a billion orbits, these orbits will stay as stable. This method is hundreds of thousands of times quicker than the way it is performed daily.
SPOCK will rule out orbits that end in a disaster even though it can not tell you how it looks like an alien solar system. This will help astronomers narrow down will distant exoplanets to study, and in the future, we'll be able to get more accurate data on exoplanet orbits. This is in the future, however, so, for now, we have to focus on the AI.