It looks like researchers that are operating with the Kepler Space Telescope have finally gained access to one Google’s AIs made for discovering exoplanets, a model that NASA worked with. What is this model called? The model, or TensorFlow, as it has been named, was used to help decode the large amount of data collected by the Kepler Space Telescope. The data managed to uncover two new exoplanets, out of which one looks to have the same eight planet system that we also have.
The new planets
The team of scientists that worked on this decided to name one of the discovered planets Kepler-90i. This planet orbits its star every 14.4 days and scientist found that it is a rocky type of planet. When talking about size, this one is about 30 percent bigger than Earth and it is even unbearably hot, temperatures ranging at about 426 degrees Celsius. The other one that was discovered was named Kepler-90, a planet that is approximately 2,500 light years away from Earth. This one is a star that strongly resembles our own Sun.
The Google AI and how it was used
One of the Software Engineer in charge of the AI model stated that he is excited to release the code that they used in order to analyze the data so that others can look at it as well. So far, they only managed to loo at 670 stars out of the massive amount of data that Kepler collected, data that refers to about 200 thousand new planets.
A computer was trained by scientist to decode all this data and the machine is the one responsible for making this interesting discovery. The computer created a neural network that it used to order through all the data collected and find the most interesting pieces of information. However, there is still more of it to go through