Bright Transient Survey Bot (BTSbot), an artificial intelligence-based algorithm created by scientists from Northwestern University in Chicago, USA, has independently found, confirmed and classified a supernova for the first time, then published data on the object. This situation was officially reported Web site scientific institution.
Over the past six years, people have spent nearly 2,200 hours visually examining and classifying supernova candidates. The new tool will help researchers accelerate discovery efforts.
The researchers trained the neural network using more than 1.4 million historical images of various classes of supernovae, galaxies, and stars from nearly 16,000 sources.
To test BTSbot, scientists turned to the recently discovered supernova candidate called SN2023tyk.
The algorithm discovered the object while scanning data from the ZTF robotic observatory. The neural network requested additional information from the Palomar Observatory in California, which houses another robotic telescope, the SED Machine. The SED then sent information about the supernova spectrum to Caltech.
After processing the dataset, BTSbot identified SN2023tyk as a Type Ia supernova, resulting from the explosion of a white dwarf.
“For the first time in history, a team of robots and AI algorithms discovered and identified a space object and then contacted another telescope to confirm the discovery of a supernova,” said Adam Miller, who led the study.
Earlier astronomers to create More than 500 thousand undiscovered stars in our Milky Way galaxy.