After AI Training, Artificial Intelligence Discovers a Supernova Without Human Intervention
For the first time, an artificial intelligence system has been able to find, confirm, and classify a supernova without human intervention. This achievement was made possible by the Bright Transient Survey Bot (BTSbot), developed by an international team of scientists, as reported by Science Alert.
BTSbot identified a supernova named SN2023tyk. It processed data from the Zwicky Transient Facility (ZTF), a robotic camera in California that scans the northern sky every two days. ZTF captured a cosmic explosion in the night sky on October 3, and BTSbot found the supernova in the data received from ZTF on October 5. After cross-referencing the information with other automated tools, BTSbot was able to confirm the discovery and classify the event as a Type Ia supernova, publicly sharing the report on October 7.
"ZTF has been operating for six years, and during that time, I and other researchers have spent over 2,000 hours visually inspecting supernova candidates and determining which ones can be observed through spectroscopy. BTSbot will eliminate the need for us to spend time verifying these candidates in our workflow," said astronomer Christopher Fremling from the California Institute of Technology.
BTSbot's AI training was conducted using over 1.4 million astronomical images from nearly 16,000 sources. Northwestern University, one of the oldest universities in Illinois, reports that the new system allows the automation of the entire process of detecting star explosions, not only eliminating human error but also significantly increasing work efficiency.
"Removing humans from the process gives the research team more time to analyze their observations and develop new hypotheses to explain the origins of the cosmic explosions we observe," said astronomer Adam Miller from Northwestern University, one of the leading researchers involved in BTSbot's development.
Although supernovae are bright and high-energy events, they are not very common and can be challenging to detect. Traditional methods of searching for supernovae rely on astronomers visually inspecting large volumes of data from robotic telescopes that continuously scan the night sky in search of new sources of light. The use of AI is expected to greatly simplify this process.
It's worth noting that astronomers from the Gaia observatory of the European Space Agency conducted new research on the Milky Way and the Solar System using the Gaia space telescope. Through their observations, they discovered 500,000 new stars in the galaxy and identified 156,000 new asteroids in our Solar System.