NASA, Google use neural network to find new planets

NASA and Google have applied a neural network to archived data from the Kepler space telescope and found “the first known eight-planet system outside of our own”.

The neural model runs using Google-developed TensorFlow technology and was trained on a dataset of 15,000 - out of a possible 35,000 - suspected planetary “signals”.

These signals are in the light readings recorded by Kepler - what NASA says are “minuscule changes in brightness captured when a planet passed in front of, or transited, a star".

“The measured brightness of a star decreases ever so slightly when an orbiting planet blocks some of the light,” Google said.

“The Kepler space telescope observed the brightness of 200,000 stars for 4 years to hunt for these characteristic signals caused by transiting planets.”

The TensorFlow model was trained to “detect the pattern of a transiting exoplanet”, and when the model was shown to work on a new sample of data, “the researchers directed their model to search for weaker signals in 670 star systems that already had multiple known planets".

“Their assumption was that multiple-planet systems would be the best places to look for more exoplanets,” NASA said.

The researchers found two planets that had been missed in previous scans of the data.

The fact they had been missed before was not hugely surprising.

“It’s like sifting through rocks to find jewels,” said Andrew Vanderburg, a NASA Sagan postdoctoral fellow and astronomer at the University of Texas.

“If you have a finer sieve then you will catch more rocks but you might catch more jewels, as well.”

One of those planets - Kepler 90i - is gaining the most attention.

“Kepler 90i is the eighth planet discovered orbiting the Kepler 90 star, making it the first known eight-planet system outside of our own,” Google AI senior software engineer Chris Shallue said.

However, there are some key differences with our own eight-planet system.

“In this system, planets orbit closer to their star, and Kepler 90i orbits once every 14 days,” Shallue said.

“The Kepler-90 star system is like a mini version of our solar system. You have small planets inside and big planets outside, but everything is scrunched in much closer,” Vanderburg added.

NASA also noted that “other planetary systems probably hold more promise for life than Kepler-90".

“About 30 percent larger than Earth, Kepler-90i is so close to its star that its average surface temperature is believed to exceed 800 degrees Fahrenheit (426.6 degrees Celsius), on par with Mercury,” the space agency said.

“Its outermost planet, Kepler-90h, orbits at a similar distance to its star as Earth does to the sun.”

The researchers now plan to apply the model to the entire Kepler dataset.

“We’ve only used our model to search 670 stars out of 200,000,” Shallue said.

“There may be many exoplanets still unfound in Kepler data, and new ideas and techniques like machine learning will help fuel celestial discoveries for many years to come.”