Artificial intelligence has discovered an eighth exoplanet orbiting a Sun-like star 2,545 lightyears away, making it the first system to be discovered with as many planets as our Solar System. Kepler-90i is a hot, rocky planet that orbits its star once every 14.4 days.
It is about 30 per cent larger than Earth, and orbits so close to the star that its average surface temperature probably exceeds 400°C.
The exoplanet was discovered using ‘machine learning’ developed by Google to search through data collected by NASA’s Kepler Space Telescope.
Machine learning is a technique through which computer programs can ‘learn’ as they operate.
The program that was used to find Kepler-90i was able to identify planets by searching through data and locating instances in which the telescope recorded dips in starlight.
These dips can indicate a planet crossing in front of – or ‘transiting’ – a star.
Exoplanet researchers Christopher Shallue and Andrew Vanderburg trained a computer to learn how to spot exoplanets using this technique.
It was able to spot weak transit signals from the eighth planet.
The Kepler Space Telescope has been operating for about four years, and in that time has built up a dataset containing 35,000 planetary candidates.
As the technology to capture large datasets increases, artificial intelligence could become key in analysing masses of data much quicker than the human brain can.
“These results demonstrate the enduring value of Kepler’s mission,” says Jessie Dotson, Kepler’s project scientist at NASA’s Ames Research Center.
“New ways of looking at the data – such as this early-stage research to apply machine learning algorithms – promise 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.”
To learn more about how AI is revolutionising astronomy, watch out for the January issue of BBC Sky at Night Magazine, on sale 21 December.