Drones are getting stealthier, and detection technology is struggling to keep up. James Purtill* says scientists have turned to hoverfly brains for help.
A Russian-made drone reportedly carrying an “air bomb” flew at least 560 kilometres over Croatia, Hungary and Romania — all members of NATO — entirely undetected by air defences.
It was just the latest urgent reminder that as drones get more stealthy, drone detection technology is struggling to keep up.
At the same time, the war in Ukraine has shown us the power of drones, visible in the grainy birds-eye videos of missile strikes on Russian tanks.
In recent years, cheap and easily available drones have shut down entire airports and even been used in assassination attempts on national leaders.
This has fired global interest in new and better ways to detect these tiny flying machines.
Now researchers from University of South Australia and Flinders University are reporting a marked improvement in detection ranges, thanks to an unlikely source: the hoverfly.
By painstakingly measuring and modeling the neurology of the hoverfly’s vision system, they’ve built an algorithm that extends detection ranges by up to 50 per cent, the researchers say.
The findings were published in The Journal of the Acoustical Society of America.
How do you map a hoverfly’s brain?
Very slowly, and with a very sharp electrode.
For well over a decade, Russell Brinkworth, an expert in autonomous systems at Flinders University, has been assembling a model of the hoverfly’s brain.
But why the hoverfly?
“Flies are much smaller and less complicated than people,” he said.
“Also, flies are really good at flying.
“Flies are so good at flying, we call them flies.
“So if you want to know how to fly, looking at a fly for inspiration is a good idea.”
Flies’ aeronautical flair partly rests on their rapid visual processing of the world around them.
Though their brains are small, each one still has a million neurons.
And these neurons are in turn tiny — smaller than can be resolved under a light microscope.
“They’re smaller than you could possibly see,” Dr Brinkworth said.
To map the brain, the researchers used a probe that had to be even smaller than a neuron, and could only record the activity of one neuron at a time.
“And so you have to hit an invisible target with an invisible spear tip.
“It’s very difficult.”
Over many years, Dr Brinkworth and his colleagues shined lights in flies’ eyes and recorded the response of individual neurons.
Eventually, they had a mathematical model of how hoverfly brains process visual signals, said Anthony Finn, an expert on sensor processing at the University of South Australia.
“Essentially what we’re talking about here is an algorithm,” he said.
How to find a drone 4km away
This bio-inspired model proved to be better than existing ones at processing and making sense of visual signals, from resolving glare and finding objects in scenes that are very bright or very dark, to discerning targets against cluttered backgrounds, Professor Finn said.
The next step was to see if the visual model worked with acoustic signals.
“We took the signal processing concepts of the vision system and then applied them to the acoustic area,” he said.
“Instead of processing, say, an individual pixel, you are processing an individual frequency channel.”
The algorithm worked just as well on acoustic signals.
“That was really surprising, even to us who thought it should make a difference,” Dr Brinkworth said.
“I thought it would be effective, but I wasn’t dreaming it would be quite as effective as it was: a 50 per cent improvement.”
In one instance, they were able to detect and locate a drone 4km away.
They saw the same 50 per cent improvement when using the algorithm to process visual and infrared signals.
Reece Clothier, a drone expert who was not involved in the research, said the approach showed promise.
“There is significant global interest in systems that can reliably detect drones at sufficient range to make intervention or mitigation possible,” he said.
“To be deployed more widely, it would have to undergo further testing, he added.
Dr Clothier, who’s the president of the Australian Association for Unmanned Systems and advises the Civil Aviation Safety Authority (CASA) on drone safety, said this would include testing in places with higher background noise, as well as visual and acoustic obstructions.
“The system was tested in Woomera, which is very benign environment,” he said.
“This is entirely appropriate for early stage development, testing and evaluation.
“The challenge will come when testing the system under more realistic conditions.”
Why the focus on detecting drones?
The research is co-authored by the Sydney-based defence company Midspar Systems and part-funded by the Australian Defence Force.
“Because the ADF are really interested in it, doesn’t mean that it only has military applications, but it definitely does,” Dr Brinkworth said.
“There are lots of reasons why you want to know where drones are.”
In 2018, a single drone shut down London’s Gatwick airport for 33 hours, disrupting more than 1,000 flights and affecting 140,000 passengers.
The drone was never captured, nor the identity of its operator ever revealed.
That same year, two drones laden with explosives were used in a foiled assassination attempt on Venezuelan President Nicolas Maduro.
One of the drones crashed into the side of a building, while the other exploded over uniformed guardsmen and caused injuries.
The two events were a wake-up call for the covert power of drones.
In the years since, governments, private airport operators, militaries and other organisations have rushed to boost their detection capabilities.
Australian company DroneShield, for instance, sells detection systems around the world, including to the US government and military.
In 2021, it reported a 91 per cent increase in revenue to more than $10 million.
CASA and Airservices Australia are developing a national drone detection system for 29 airports around the country.
Russia’s invasion of Ukraine has put drones in the spotlight once again.
Ukraine’s small fleet of Turkish-made Bayraktar TB2 armed drones, which cost about $US1 million each, have destroyed vast numbers of Russian vehicles.
Ukraine has been using drones to “great effect”, a Pentagon official said last week.
Dr Brinkworth said the technology could also be used for better crash-detection systems in driverless cars, or to improve cochlear implants and boost the microphones on mobile phones.
“You could have a whole mess of people talking at the same time and it would help to differentiate the background from the foreground sound,” he said.
“It just goes to show how the adaptation that goes on in biology is so well tuned and so universally applicable.”
*James Purtill works at triple j Hack in Sydney.
This article first appeared at abc.net.au.
