- Fiza Pirani The Atlanta Journal-Constitution
Looks like our solar system isn’t the only one out there with eight planets circling around a single star.
NASA scientists, along with Google engineers, used artificial intelligence to discover a new, scorching hot planet in the Kepler-90 solar system, bringing the total number of planets circling the Sun-like star to eight, the agency announced Thursday.
Scientists used Google machine learning to teach computers how to identify planets in the light readings recorded by NASA’s planet-hunting Kepler space telescope.
Here’s what you should know about the new discovery, according to Thursday’s live teleconference:
What exactly is the Kepler mission?
The Kepler mission, NASA Discovery’s tenth mission, first launched in March 2009 with a goal to survey the Milky Way and hunt for Earth-size and smaller planets near the galaxy or “habitable” regions of planets’ parent stars.
In 2014, the Kepler space telescope began a new extended mission called K2, which continues the hunt for planets outside our solar system along with its other cosmic tasks.
According to Space.com, “Kepler spots alien worlds by noticing the tiny brightness dips they cause when they cross the face of their host star from the spacecraft's perspective.”
Since 2009, Kepler has discovered thousands of exoplanets ranging between Earth-size and Neptune-size (four times the size of Earth).
As of Dec. 14, Kepler has confirmed 2,341 exoplanets.
How did NASA find the planet?
Researchers Christopher Shallue (Google AI software engineer) and Andrew Vanderburg (NASA astronomer) were inspired by the way neurons in the human brain connect and adapted the “neural network” concept to machine learning.
They taught computers how to identify planets in the light readings recorded by the Kepler telescope by first training them to search for the weaker signals in 670 star systems that already had multiple planets.
“Their assumption was that multiple-planet systems would be the best places to look for more exoplanets,” researchers wrote in the press release.
Using this concept, the network “found weak transit signals from a previously-missed eighth planet orbiting Kepler-90, in the constellation Draco.”
Their findings will be published in The Astronomical Journal.
Would humans have found the planet without machine learning?
Without machine learning, it would have taken humans much longer to scan the recorded signals from planets beyond our solar system (exoplanets), Shallue said. Kepler’s four-year dataset consists of 35,000 possible planetary signals.
Additionally, people are likely to miss the weaker signals that machine learning was able to identify.
Won’t this form of automation put astronomers out of work?
"This will absolutely work alongside astronomers," Jessie Dotson, Kepler’s project scientist at NASA’s Ames Research Center in California’s Silicon Valley, said in a press briefing. "You're never going to take that piece out."
Researchers hope astronomers will use this form of automation via machine learning as a tool to help astronomers make more of an impact, increase their productivity and inspire more people become astronomers.
What is the new planet like?
The new “sizzling hot, rocky planet,” named Kepler-90i, appears to be the third planet from its Sun-like star.
"Kepler-90i is not a place I'd like to go visit," Vanderburg said during a press briefing.
It measures approximately 800 degrees Fahrenheit and doesn’t have a very thick atmosphere, according to the researchers.
That makes it unlikely that life as we know it could exist there.
It orbits its star, Kepler-90, once every 14.4 days and is the smallest planet of the bunch.
All of the planets, including Kepler-90i, are closer to their star than the Earth is to the sun.
Kepler-90i is 2,545 light years from Earth.
Are there more 8-planet solar systems out there?
According to Vanderburg, this won’t be the last time they discover a solar system with eight planets.
“Is an 8-planet system really that extraordinary? Maybe there are systems out there that make our 8-system seem ordinary,” he said.
Did the scientists find anything else using neural network machine learning?
Yes. According to NASA, they found a sixth planet — the Earth-sized Kepler-80g.
Vanderburg and Shallue plan to apply the same method to Kepler’s full set of more than 150,000 stars.
“These results demonstrate the enduring value of Kepler’s mission,” Dotson said. “New ways of looking at the data – such as this early-stage research to apply machine learning algorithms – promises to continue to yield significant advances in our understanding of planetary systems around other stars. I’m sure there are more firsts in the data waiting for people to find them.”View full experience