Arkeus: From drowning in drone data to $100 million valuation
AI Analysis
Australian defense-tech startup Arkeus has secured $25 million in Series A funding to scale production of its AI-powered hyperspectral sensing systems. This technology, capable of detecting targets at significantly longer ranges in degraded conditions, addresses a critical need for improved drone-based intelligence gathering. Arkeus is already deployed with the Australian Department of Defence and US agencies, including at the Mexican border.
Key Takeaways
- Arkeus raised $25M in Series A funding, valuing the company at $100M.
- The company’s core technology is a miniaturized “hyperspectral optical radar” offering 8x the detection range of existing systems in challenging conditions.
- Arkeus’ system reduces false positives in drone imagery analysis, a significant problem for operators.
- The technology utilizes a broad spectrum of electromagnetic signals (UV to infrared) combined with AI for rapid data analysis.
- Current deployments include the Australian Department of Defence and US border security (Mexican border).
Why It Matters
This technology directly addresses a key limitation in current drone operations – the ability to reliably detect and identify targets in adverse weather or obscured environments. Improved sensor capabilities will enhance situational awareness and effectiveness for both military and civil defense applications, potentially impacting counter-UAS operations and border security. The rapid processing time is a significant advantage, reducing operator workload and improving response times.
Arkeus: From drowning in drone data to $100 million valuation
Melbourne defence-tech start-up Arkeus has raised $25 million on its technology born from a founder seeing the frustration of drone operators hunting drugs in the Colombian jungle.
Arkeus founders Simon Olsen, left, and Dr Jonathan Nebauer
Key Takeaways
- Arkeus has raised $25 million in a Series A round led by QIC to scale its AI-powered “eyes and brain” sensing systems.
- The Melbourne-founded deftech company says its “hyperspectral optical radar” can detect targets up to eight times further than existing military optical systems in degraded conditions.
- Arkeus plans to use the money to manufacture in Queensland and the US.
- Its valuation has grown more than sevenfold to $100 million since a seed round two years ago, Arkeus says.
- The raise saw participation from new investors R+VC, Folklore Ventures and DYNE Ventures, with continued support from existing investors Main Sequence Ventures, Salus Ventures and Beaten Zone.
- The company styles its “eyes and brain” technology for autonomous craft as being able to see more and act faster on that information.
- It is deployed with the Australian Department of Defence and various U.S. defence and civil defence agencies, including at the Mexican border.
Key Background
Simon Olsen was in the jungles of La Macarena, Colombia, where the Andes and Amazon meet, looking for drug operations, when the realisation hit.
He was sitting behind US Air Force drone operators who were trying to see through the canopy of mist and foliage. “Because it was such a challenging environment, the software was generating hundreds, if not thousands, of false positives,” Olsen tells Forbes Australia. “The software thinks it sees something, but it hasn’t. And the operator spent hours – I kid you not – hours sifting through all this garbage coming through from the software system.”
An Arkeus unit at work.
It was 2016 and Olsen was a business development director for the Melbourne-based defence-tech success, Sentient Vision Systems, specialising in computer vision for the def-tech sector.
“Everyone was talking software, but we saw that software couldn’t find what the sensors can’t see. Most sensors – whether that be cameras or radars – on drones today are the same cameras and radars that have been on crude aviation assets for decades. All we’ve done is miniaturise them.”
Smoke, haze, dust, and humidity were as big a problem as ever.
Olsen’s colleague at Sentient, aviation engineer Dr Jonathan Nebauer, thought he had a solution. By using signals from across the electromagnetic spectrum – from ultraviolet through the infrared – they’d be able to cut through all the dust, smoke and haze.
And AI would then analyse it all in seconds.
Hyperspectral sensors typically were half the size of a boardroom table and the data they produced took a day to process, Olsen says. He didn’t think it could be reduced to a 4kg cube that could analyse the data