NASA | Solar System | Artificial Intelligence | Kepler spacecraft | Exoplanet | Kepler-90
The Kepler Space Telescope has helped NASA discovered thousands of different exoplanets including seven planets in Kepler-90's orbit. On Thursday 14th Dec, the space agency along with Google announced the discovery of an eight exoplanet through the help of machine learning, which now makes it the first 8-planet system outside our own. Google's AI researchers trained a computer to learn how to find exoplanet by feeding it with data recorded by Kepler.
Google AI researcher Christopher Shallue and Andrew, an astrophysicist from UT Austin, worked together to teach a machine learning system how to analyse a trove of data collected by Kepler. The researchers created a TensorFlow model to distinguish planets from non-planets using a dataset of more than 15,000 Kepler signals. They thought the machine to recognise and distinguish patterns caused by actual planets and patterns caused by starspots and binary stars. They found that the computer was able to identify which signals were of plants 96 per cent of the times.
"Just as we expected, there are exciting discoveries lurking in our archived Kepler data, waiting for the right tool or technology to unearth them," said by Paul Hertz, director of NASA's Astrophysics Division in Washington. "This finding shows that our data will be a treasure trove available to innovative researchers for years in future and comming."
Using machine learning, NASA and Google were able to discover two additional exoplanets circling Kepler-90 - one named Kepler-90i and the other Kepler-80g. The Kepler 90i is a planet that orbits its star ever 14.4 days. It is also 30 per cent larger than Earth and has an average surface temperature of over 800 degrees Fahrenheit, similar to Mercury. So essentially this looks like a pretty uninhabitable planet for us, so let's keep looking.
"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," said Vanderburg, a NASA Sagan Postdoctoral Fellow and astronomer at the University of Texas at Austin.
Shallue's idea to apply a neural network to Kepler data came about during Google's 20 per cent project, wherein the company requires its employees to spend 20 per cent of their time for creative ideas. "In my free time, I started googling for 'finding exoplanets with large data sets' and found out about the Kepler mission and the huge data set available," said Shallue. "Machine learning really shines in situations where there is so much data that humans can't search it for themselves."
Google in recent years has been investing heavily on AI and machine learning across its products and services. The latest discovery is just one of the few breakthroughs in machine learning in recent times including work done by Google's DeepMind, such as AlphaGo.