Two exoplanets previously unknown to astronomers have been found using artificial intelligence. A group of young researchers utilized AI from Google to analyze data from the Kepler mission, uncovering these two worlds. One planet, K2–293b, is found 1,300 light years from Earth, while K2–294b is located 1,230 light years away from our own system. Both planets are seen in the constellation Aquarius.
The Kepler space telescope launched in 2009 on a mission to discover new worlds orbiting alien suns. Following the failure of gyroscopes used to orientate the observatory, the program entered the K2 segment of its mission. It was during this time that the spacecraft collected the observations which contained data on the newly-discovered worlds.
“[The two planets] are both very typical of planets found in K2. They’re really close in to their host star, they have short orbital periods, and they’re hot. They are slightly larger than Earth,” explained Anne Dattilo, an undergraduate at the University of Texas at Austin, leader of the team that made the discovery.
Dattilo developed the algorithm used to reveal the unknown planets in the Kepler data, working with Andrew Vanderburg from UT Austin and Google’s Christopher Shallue.
Vanderburg and Shallue were part of the team which first utilized AI to reveal an exoplanet in observations from Kepler in 2017. That exoplanet was discovered in Kepler-90, a system already known to possess seven planets, tying that family of worlds with our own for the largest known number of planets in one solar system.
“For the first time since our solar system planets were discovered thousands of years ago, we know for sure that our solar system is not the sole record holder for the most planets,” Vanderburg said in 2017.
Once AI had revealed evidence of the exoplanets in the Kepler data, the team studied the host stars in visible light, utilizing the the Whipple Observatory in Arizona and the Gillett Telescope in Hawaii. These observations confirmed the presence of these newly-discovered worlds.
Following the failure of gyroscopes ending the original Kepler mission, the spacecraft wobbled, a movement which had to be accounted for during the K2 segment of the program. After the discovery of thousands of worlds during the original phase of its mission, astronomers found hundreds more during the K2 segment. Today, nearly 4,000 exoplanets have been found by astronomers, with a similar number of findings awaiting confirmation.
Artificial intelligence could prove vital in the search for exoplanets seen by observatories like Kepler, that may not otherwise be seen through conventional methods.
“Even if every star had an Earth-sized planet around it, when we look with Kepler, we won’t find all of them. That’s just because some of the data’s too noisy, or sometimes the planets are just not aligned right. So, we have to correct for the ones we missed. We know there are a lot of planets out there that we don’t see for those reasons,” Vanderburg explained.
NASA ended the K2 mission, and with it the Kepler program, in October 2018. However, the search for exoplanets continues with the Transiting Exoplanet Survey Satellite (TESS) observatory, launched in April 2018.
The techniques developed by the team, including the algorithm created by Dattilo, could be used by other researchers searching for additional planets hiding in Kepler data. With 300,000 stars examined by Kepler, and a wealth of data set to arrive from TESS, artificial intelligence is going to be doing a lot of planet hunting in the very near future.