Flare Bright CEO Kelvin Hamilton talks to Editor-in-Chief Andrew Drwiega about GPS-free navigation for uncrewed aerial systems.
What was your career path into Flare Bright?
I did a degree in Cardiff, and got interested in autonomous systems robotics. So I ended up doing a PhD at Edinburgh in subsea robotics, and co-founded a company called Seebyte, which became quite successful in the subsea space.
We were addressing subsea autonomy and running without any GPS, without any pilot communications. When we sold that company in 2013, I started an aerospace company taking highly autonomous systems into the airspace where there wasn’t much autonomy at the time.
Eventually I had an idea for a tiny Nano Drone for reconnaissance. It was originally for search and rescue, but it was hard to sell into that field. But the British Army were interested in it for tasks such as soldier reconnaissance. Of course the little drone deliberately didn’t have GPS in it, just because you don’t really need it with the short mission being ‘what’s behind the building’ or ‘around the next corner.’
So we designed out the GPS. We self funded for a couple of years and then we got investment for around half a million pounds just when the COVID lockdown was starting.
What was the crucial break for you?
We got two opportunities with the government. We got invited to participate in the U.K. government’s Defence and Security Accelerator (DASA) in late 2020, then received Future Flight grants of over £400,000 from Innovate UK at the start of 2021, which then accelerated the investment due to the credibility it brings with it.
Why was there an interest in GPS-free guidance? The invasion of Ukraine had not happened and the panacea was that GPS was the way to go.
Initially we designed GPS out because it was this little drone which only flew for around for a very short time, and soldiers wanted to use it instantly. Imagine the original application for the search and rescue mission. If someone fell overboard, someone grabs a pistol and fires a smart flare. It does a scan to locate the person then provides that information. So it didn’t need GPS. You don’t need to wait a few minutes for the satellites to lock because it then stops it being an instant system.
In 2019 we exhibited at the SOFINS Special Operations Forces event in France. There were many drone demonstrations but many had problems working due to the presence of drone jammers and radios. But ours worked and people noticed we were GPS free. So we got noticed. We had the DASA funding and our tiny nano drone. So we were approached by people who had loads of drones, but no GPS independent systems. We were asked if we could put that capability onto bigger and bigger platforms and show that it works just as well.
How easy is it to scale it up?
It’s very easy. In fact, the smaller the drone, the harder it is. So we’ve been going from what started off with a 10 centimetre wing size (the width of a tennis ball), and now we’re up to drones with a five metre wing span – and there’s no issue in weight.
And as soon as Ukraine got invaded, suddenly GPS free became a really big deal. So now we’re getting onto bigger and more operational platforms and improving the performance of the system in parallel.
Which other organisations have shown interest in what you do?
The U.S. Army Combat Capabilities Development Command (DEVCOM). Their interest came through DASA because we participated in a big DEVCOM demonstration day. They didn’t really care about the nano drone, but really liked the GPS free. We showed our minimum viable product, and it did its job. It showed what the market really wanted, and now we focus on that.
What was the result?
Well nobody wanted the hardware, but they wanted the software. So that’s okay for us, because really, we’re a software company at heart – we’re a bit like Intel inside. We make this software on the drone already and gives it a massively improved GPS free navigation capability.
Where did you go from there?
Last year, we were receiving parallel funding from Ministry of Defence (MoD) and U.S. Department of Defence (DOD). We deployed to the Mediterranean to show that it also works over water, as many UAVs cannot fly over the sea without GPS. Flights of up to one hour were conducted over land and sea to prove the capability of the software to continuously analyse the speed and position of the UAV using data and signals of opportunity, resulting in very accurate ‘fixes’ on its location.
It worked really well, and so now we’re in conversations regarding how to take this capability to the operational level, but obviously it all takes time.
How does your GPS-free navigation system work?
It’s really a dead reckoning system. Think about navigation in an aircraft, such as a World War Two bomber. They navigator knows the speed of the plane and its performance characteristics and external factors such as the weather and wind direction through the mission. Basically a dead reckoning system without GPS. That’s basically what we do, but we do it microsecond by microsecond, instead of every 10 minutes. The system is loaded with known data about the aircraft/drone, then it receives data from the aircraft such as the propeller revolutions per minute (rpm) and the angles of the control surfaces which is all fed back into the analysis part which is housed in a small computing tablet.
So all of that information goes into basically dead reckoning. We pull in every bit of information we can, and that just tells you, I’ve flown, you know, three miles on this bearing, but with great accuracy as it is constantly updating itself.
When you create a duplicate of the platform, how does that inform what you are doing?
We model it and call it a digital twin. We make a highly accurate model – a digital twin – of the aircraft, and use that with other modelling and data and forecasts. The data is fused which allows us to predict of where the platform is and how it is moving. And then that ends up being better than the competition.
Can you use this on bigger UAVs such as a Medium Altitude Long Endurance (MALE) aircraft?
What we find is that, with a bigger platforms, they usually just fit an inertial navigation system. Where we come in is the really small, cheap, attritable systems, where they can only fit a $5 chip. It’s a an extremely low cost addition to something they’ve already got. All they have to do is fit software. And If their GPS is denied, it acts as a back-up.
Is it all software driven without any external input from any sensor?
Well, we use every sensor, all the ones on the larger fixed wing UAVs that are already present for the mission. We use the air speed indicator, the air temperature sensor as well as the humidity and propulsion sensors. In small UAVs/drones we use the command signals as well as all the digital twin modelling we have done on each platform.
What are your plans going forward?
We have around 25 people in the company at the moment we need more as we are growing in every direction. We are looking at opening international offices, particularly in the Middle East and Asia. We will be ISO 9001 compliant very soon, which shows that we’ve got the business processes, and then we just start to expand.
We are also getting involved in rapid testing and evaluation using machine learning and AI technology. This will allow us to develop our software with a higher capability – and faster.
Thank you.
