Blog icon

23 March 2023 4 min read

From analysing sound waves within stars to mapping galaxies, AI is already playing a pivotal role in space exploration. It's even helping space explorers land rovers autonomously on nearby planets. We spoke to leading astrophysicists, astronomers and space engineers in the field to find out more.

This blog is an excerpt from episode six of our Everyday AI podcast. this episode of Everyday AI, we look up and out to the stars with astrophysicists, astronomers and space engineers.

Gazing into our past

For Wiradjuri astrophysicist and science communicator, Kirsten Banks, understanding the sky above is a key part of understanding our past. Kirsten is currently in a PhD program, studying the stars of our Milky Way galaxy to try and understand more about its history and formation.

“The really cool part about being a Wiradjuri astrophysicist is that you do into space to understand more and reveal more knowledge,” Kirsten said.

“But there's also that history behind you, with that culture that's been existing on this land that we now call Australia for over 65,000 years."

Kirsten’s research involves studying galaxies to understand their evolution over the past 13 billion years. Her work is looking specifically at two very similar types of stars: Red Giant Branch stars and Red Clump Stars. To identify these stars, Kirsten is looking very closely into their DNA.

“These observations are very time intensive with many observations.They take weeks, months, sometimes even years," Kirsten said.

This is where AI comes in. Kirsten is using a data-driven algorithm called the Cannon. This algorithm learns from the data of these stars and generates a model based on parameters fed into it. This saves an enormous amount of work for Kirsten and her colleagues. It allows to them to more easily tell whether they’re looking at the Red Giant stars, or the Red Clump stars. Researchers can use this information to map the galaxy, understand how old the stars are, and how they’ve travelled throughout the galaxy.

"So if we know exactly where they are, where they've been, and how old they are, and what they are made of, we can rewind time," Kirsten said.

"And through that, understand and try and answer more questions like, where did life come from?”

Black hole breakthroughs

When you look up at the stars, what do you see? According to our astronomer Dr Ivy Wong, the answer is mostly hydrogen and helium, also known as baryonic matter. All that matter combined with dark matter is what makes a galaxy.

Ivy's research involves figuring out how galaxies form stars and grow central supermassive black holes. Her work has recently enlisted the help of AI to try to automate ways to classify galaxies.

A blurry orange ring of light against a dark background

As astronomer's instruments get more advanced, we end up with bigger and bigger data sets. Terabytes of data. An upcoming survey of star-forming galaxies called EMU will use our newest radio telescope, ASKAP, to look deep into space. This survey is expected to detect up to 40 million galaxies across the history of the universe. Many of these will be millions of light years away from us. Some will show us black holes. We may even detect new objects we've never seen before. Our ASKAP radio telescope is recording about 75,000 terabytes of raw data per year.

With all this information, astronomers like Ivy will be able to observe galaxies at their different stages of evolution, to study how they change as they evolve.But it’s a big job. That’s where AI comes in.

“AI computer vision is at a stage where it can analyse these galaxy images and sort them into their varying stages. But it’s only good enough to sort 90 per cent of the 40 million," Ivy said.

"The remaining four million galaxies are left to human eyeballs. That’s still too big a task.”


Space-age technology

Over a few years from 2013-2019, Ivy’s team did a call out to everyday people to help them. You could log into one of these projects and review pictures to identify specific galaxies. This process created one of the best quality data sets possible as it had been reviewed by multiple people. Experts then used the dataset to train the AI so it could analyse the final four million galaxies. And leave a more manageable number for a team of citizen scientists.

The AI system called CLARAN (CLAssifying Radio sources Automatically with Neural network) uses the same technology that Facebook uses to detect people's faces in your photos. Essentially learning to recognise certain features and suggest who or what we’re looking at.

"CLARAN would draw a box and collate together the different blobs into what it thinks is one object," Ivy said. "We taught CLARAN how to identify six different types of classifications of radio galaxies.”

With the help of CLARAN, our understanding of the universe continues to grow.

"The number of galaxies and types of galaxies is fixed, but our ability to see them actually increases in time," Ivy said.

"We may not be seeing all the galaxies out there at this point in time. And with more advanced technologies, we see weirder and weirder galaxies."

Contact us

Find out how we can help you and your business. Get in touch using the form below and our experts will get in contact soon!

CSIRO will handle your personal information in accordance with the Privacy Act 1988 (Cth) and our Privacy Policy.

First name must be filled in

Surname must be filled in

I am representing *

Please choose an option

Please provide a subject for the enquriy

0 / 100

We'll need to know what you want to contact us about so we can give you an answer

0 / 1900

You shouldn't be able to see this field. Please try again and leave the field blank.