Artificial Intelligence detects a “supernova” on its own for the first time in the world Both taking pictures – confirming – publishing without any human participation at all
Scientists spend a great deal of time every day trying to detect star deaths known as supernovae, and their insights help researchers understand more about stellar evolution.
.
Astronomers usually look for points of light that were not previously there in the sky. Then I asked the different observatories to confirm and collect the different data to classify it. They then publish (publish) their results.
.
But with the advent of “artificial intelligence” or AI, scientists hope it will help reduce the time required. So they developed a new tool called Bright Transient Survey Bot (BTSbot), an AI that learns from more than 1.4 million images from nearly 16,000 sources to look for one thing: supernovas.
.
And they succeeded in the real world on October 3rd. The “Zwicky Transient Facility” (ZTF) astronomical observation center is also an automated telescope. An image of one area of the sky is taken and stored in the data.
.
Only two days later, on October 5, BTSbot was able to detect the newly formed spot of light. It was named “SN2023tyk.”
.
But instead of just informing astronomers, the AI made contact with another robotic telescope called SEDM (Spectral Energy Distribution Machine) at Palomar Observatory. Collect spectral data to classify and verify it from other sources.
.
Once the supernova’s existence was confirmed, BTSbot published its findings on the International Astronomical Union’s website on October 7. Without any human involvement in the process, the discovery identified it as a supernova, type La, about 760 million light-years from Earth.
.
“This is the first time that a combination of robotics and artificial intelligence algorithms has been observed and then identified and contacted with another telescope to confirm the discovery of the supernova,” said Adam Miller, head of the Artificial Intelligence Project at Northwestern University.
.
In the case of an LA-type supernova, it is an explosion from a white dwarf that sucks in the material of its companion star.
.
When a large amount of matter flows in, the white dwarf will burn up and explode, destroying them both. It creates light that may be brighter than all the stars surrounding it combined, yet it is very difficult for humans to notice.
.
“We have achieved the world’s first fully automatic detection, identification and classification of supernova,” added Nabil Rahimullah, a scientist on the AI development team.
.
“This helps provide more data to study large supernovae. It helps us better understand the life cycle of stars and the origins of the elements that supernovae create, such as carbon, iron and gold.
.
These artificial intelligences will allow scientists to spend time on other analyses. Make the work of astronomy more efficient and let artificial intelligence do its work
.
“When everything is working properly, we’re not really doing anything,” Rahimtola said. We sleep at night and in the morning we see BTSbot and other AI systems continuing to do their work.”
.
This system was developed by researchers from several institutions including: Northwestern University and is collaborating with astronomers at the California Institute of Technology, the University of Minnesota, Liverpool John Morris University, and Stockholm University.
Researched and edited by Witit Borompichaichartkul
Image credit: Paleosurveys/D. Lang (Oceanic Institute) for Paleosurvey stratigraphy and unWISE/NASA/JPL-Caltech/D. Lang (Peripheral Institute)
Origin
https://www.wis-tns.org/object/2023tyk/classification-cert
.
https://www.wis-tns.org/astronotes/astronote/2023-265
.
https://www.iflscience.com/supernova-detected-confirmed-and-classified-by-ai-for-the-first-time-71142
.
https://www.sciencealert.com/world-first-new-ai-system-discovers-supernova-without-human-help
.
https://www.space.com/artificial-intelligence-spots-first-supernova