This site may earn affiliate commissions from the links on this folio. Terms of use.

NASA'south Kepler satellite has been observing the stars for almost a decade, and information technology'southward produced a mountain of data in that time. Late last year, Google showed how car learning could help astronomers dig through the Kepler backlog, and it discovered a few new exoplanets in the process. Google has now open up sourced the planet-spotting AI then anyone can give it a shot.

We tin't observe exoplanets straight yet, but nosotros can spot the telltale dip in luminance when an exoplanet passes in front of its host star. So, Kepler was designed to scan big swaths of the sky in search of these signals. However, there are a substantial number of candidate signals that might point to an exoplanet. We don't know for certain until astronomers tin examine the information more closely, merely there's too much to check every reading. That'south why astronomers have simply checked the 30,000 or and then best candidates, which has resulted in the discovery of around ii,500 exoplanets.

Google'south solution to the data backlog was to unleash the power of neural networks on the problem . That seems to be Google's go-to idea lately, whether you're talking about cocky-driving cars or smartphone photography. Using fifteen,000 examples of exoplanets in the information, Google trained its network to sift through Kepler data and identify other exoplanet signals. Later on the training period, Google's AI checked 670 stars that were known to accept exoplanets. And wouldn't you lot know information technology, Google spotted two new exoplanets that would take otherwise gone unnoticed.

A neural network is skilful at recognizing patterns, just you need a lot of data to train the network. Google is saving anybody the time of training a neural network on Kepler data by releasing its code freely. You can get the TensorFlow code on GitHub right now, and Kepler's data is also freely available. Information technology'due south non as easy as grabbing both piles of lawmaking and launching an app, though. Experience with Google'southward TensorFlow platform and Python volition help. Google's weblog postal service provides instructions on how y'all tin can test the network past re-discovering Kepler 90i, one of the planets Google found in the original test.

The Googlers behind this projection promise that the customs will make employ of the code to continue the hunt for planets in Kepler's information. At that place are thousands more signals to analyze, and that could be just the commencement. Googlers hope to develop more avant-garde neural networks to dig through NASA data for Kepler and future missions like the upcoming Transiting Exoplanet Survey Satellite.