Google's AI helped NASA to find another planet in Kepler-90's orbit

Postado Dezembro 16, 2017

Google and NASA scientist claims to find two new plants in our solar system.

A solar system with as many planets as our own has been discovered with the help of NASA's Kepler space telescope and artificial intelligence, the U.S. space agency said.

A planet named as Kepler-90i in the new system is a rocky planet like earth, but it completes its rotation around its star in 14.4 days, which means Kepler-90i's day is two weeks of Earth. The data was then analysed using machine learning technology from Google.

The researchers say they plan to use the same method to search the remaining 150,000 stars the Kepler space telescope has studied so far.

The researchers taught a computer to learn how to identify the weak signals of a planet crossing its star, getting it to search for planets around 670 known multiple-planet systems. Some form of migration occurred, dragging this system inwards, producing the orbits we see in Kepler-90 today. It's in a planetary system called Kepler-90, located in the constellation Draco, some 2,545 light years from Earth.

Image: A comparison of Kepler-90 and our own solar system. "You have small planets inside and big planets outside, but everything is scrunched in much closer", said Andrew Vanderburg, astronomer and NASA Sagan Postdoctoral Fellow at The University of Texas, Austin. And it discovers new planets in existing systems, observing signals that seem to be of interest, but are too weak to be noticed by humans.

NASA has found a distant star circled by eight planets, equal to the complement in our own Solar System.

Machine learning neural networks have been assisting astronomers for a few years now.

While machine learning has previously been used in searches of the Kepler database, this research demonstrates that neural networks are a promising tool in finding some of the weakest signals of distant worlds.

"Machine learning really shines in situations where there is so much data that humans can't search it for themselves", said Shallue, a senior software engineer at Google AI and lead author of the paper accepted for publication in The Astronomical Journal. Though Kepler looked at one portion of the sky for stars that were farther away for a longer time, TESS will observe the entire sky and focus on the brightest and closest stars, each for 30 days.

"There's lots of unexplored real estate in Kepler system and it would nearly surprise me if there weren't more planets in this system", Vanderburg said.

The new planet has been dubbed Kepler-90i. To date, the spacecraft has discovered 2,525 confirmed planets.

Machine learning had not been applied to data acquired by the Kepler telescope until Mr Shallue came up with the idea. The computer creates a neural network like the human brain. Kepler-80g is an Earth-sized planet that is gravitationally locked in a resonant chain with four of its fellow planets, forcing them to orbit their star nearly as if they were all moving to the music of a highly choreographed dance. The researchers are planning to find more planets using this technique.

"We got lots of false positives of planets, but also potentially more real planets", said Vanderburg. "This finding shows that our data will be a treasure trove available to innovative researchers for years to come".