Space traffic management – what happens when things collide?

There is no such thing as space traffic management. Yet. “But now, all of a sudden, we really need it”, says Saber Astronautics CEO, Jason Held.

In just the past year, the number of active satellites has almost doubled to 7600. Estimates put that number at 40,000 by 2030. And there are hundreds of thousands of pieces of junk and debris.

“Recent cases of space debris landing in Australia recently is a case in point,” Held says. “Whose authority and responsibility is it to track that? Who determines if it’s an attack or an accident? Is it civilian? Is it military? Who has jurisdiction?”

Saber Astronautics is a space engineering organisation with branches in the United States and Australia. They are attempting to tackle the pressing issues of tracking who owns what in orbit, what it is doing, where, and why.

“There are all these questions,” says Held. “What happens when things collide by accident? That’s a problem for space traffic management. But that’s different from military space, which is about keeping people from doing naughty things on purpose.”

One problem is being able to distinguish between the two.

Outer space is not free of regulation, but it’s certainly not flush with it either.

The Australian branch of Saber Astronautics operates the Responsive Space Operations Centre (RSOC) based at Adelaide’s Lot 14 precinct and a research-and-development laboratory in Sydney. 

“We’re using all the tools we have in the RSOC and building up a cadre of people who are trained up in using them,” Held says. “We’ve got Australian staff who have been working with the US for quite some time who are now leading space domain awareness and space traffic exercises for the Pacific Region”.

The next major international exercise is in November.

The Saber Astronautics US branch operates an RSOC in Boulder, Colorado and an integrations unit working with the US Department of Defence in Colorado Springs.

The Sprint Advanced Concept Training (SACT) tests bring businesses and agencies in the US, Europe and Australia together to determine the cause and effect of live and simulated space “incidents”. And how to best respond to it.

“It started off as a US Department of Defence war game about ten years ago,” says Held. “Because commercial space is happening, it now has an unclassified side which we participate in.”

It could be a solar flare. A satellite collision. A cyber attack. Or a messy combination of events. 

“It’s kind of like a hackathon, the way it’s run,” he says. “It’s about people finding problems right away and trying new things right away. The catchphrase is: ‘It’s okay to suck – just suck less next time’.”

The Pacific component alone involves about 200 people from 20 different companies and research groups. And four different militaries.

Space traffic jam

“Some have radars. Some have telescopes. Some have software. Others have combinations of the above,” Held says. “It’s got big companies and little companies. You’ve got global primes working side-by-side with crispy little startups just trying to make their way.”

The problem is data.

There’s just so much of it.

But it still doesn’t cover everything, everywhere, all the time.

And it all comes in different formats, at different rates, with differing degrees of reliability.

That’s why, for example, we rarely know the exact location of any given satellite.

It may get an occasional radar fix. But, unless it’s openly broadcasting its position the way terrestrial aircraft and shipping do, its orbit is often a matter of probability mathematics.

And it’s not just a matter of tracking satellites.

“Anyone can look up into the sky and watch a space object”, Held says. “But trying to get context from that is pretty hard.”

Could space weather cause unwanted behaviour? Poor mission design? Chance? Or a malicious act? 

“The most mature piece of machine learning we have is used to figure out the cause of damage to a spacecraft,” he adds. 

It identifies the likelihood of a fault being because of a cosmic ray, solar storm, debris impact – or hack.

“You can use machine learning to detect changes in the pattern of a spacecraft’s behaviour, and that’s important – especially if you’re in the military and you’re trying to figure out if somebody is drifting towards your asset by accident or on purpose”.

The exercise aims to find the best way to prevent a disaster. No matter how ‘out there’ the idea sounds.

“We try 100 ways and see which one’s performed the best. And then everybody learns together”.