Using AI to catch the next galactic supernova in real-time
Dr Matt Edwards from Waipapa Taumata Rau – the University of Auckland will use artificial intelligence (AI) to hunt for the gravitational waves from exploding stars, making it easier than ever to detect and study these rare cosmic events
Supernovae (exploding stars) create ripples in the fabric of spacetime known as gravitational waves. Supernovae are incredibly rare, happening only once or twice a century in the Milky Way. However, current methods to analyse the data from gravitational wave detectors are too slow. It can take months to identify the signals that confirm a detection, meaning we risk missing out on insights from these cosmic explosions.
Dr Edwards has received a Marsden Fund Fast-Start Grant to develop cutting-edge AI systems to solve this problem. The team will create two powerful tools. The first is a neural network trained to generate accurate theoretical gravitational wave signals from supernovae in under a second, compared with months using current methods. The second tool automatically sifts through detector data to identify real supernova signals and instantly identify the dying star's physical properties, such as its spin and the nature of the ultra-dense matter at its core.
With this research, Dr Edwards and the team will ensure that when the next star in our galaxy explodes, we will be ready to make the most of it. The advanced AI techniques developed will help unlock secrets of stellar death and may also have broader applications beyond astronomy. For example, AI systems with rapid signal processing and pattern recognition could benefit New Zealand's tech sector in areas like financial modelling.