Scientists around the world are constantly seeking new tools and techniques to help them explore the mysteries of the cosmos. One of the most fascinating yet difficult-to-understand phenomena are supernovae, the explosions of stars that emit enormous amounts of energy and matter into space. Now, thanks to advanced artificial intelligence algorithms, astronomers from several universities around the world have made a breakthrough discovery, developing a system that independently detects and verifies supernovae, without human involvement. The system is called the Bright Transient Survey Bot (BTSbot) and already has its first discovery to its credit.
Artificial Intelligence in the Service of Space Exploration
A collaboration of scientists from Northwestern University, Caltech, the University of Minnesota, Liverpool John Moores University, and Stockholm University has resulted in the creation of a new tool that fully automates the process of searching for supernovae in the night sky. This achievement not only speeds up the analysis of new celestial objects but also helps avoid human error, which is extremely important in space research.
"For the first time in history, artificial intelligence algorithms spotted, identified, and then confirmed—by communicating with another telescope—the discovery of a supernova. This is an important step forward, and further refinement of the model will enable the algorithms to identify specific subtypes of explosions. Ultimately, removing humans from this process will give researchers more time to analyze their observations and develop new hypotheses explaining the origins of the cosmic explosions we observe." - Adam Miller, Northwestern University
The automated software in BTSbot presents scientists with a list of potential explosions, and researchers verify the candidates and conduct spectroscopic observations. A spectrum, or the distribution of light waves of different wavelengths, is key to confirming whether an object is a supernova. There are robotic telescopes capable of collecting spectra, but humans often operate telescopes with spectrographs.