Resilient navigation and positioning
On an open airfield under a clear sky, a centimetre is free. We are hired for the rest of it — the tunnel portal, the hangar roof, the steel structure, the jammed sector — where the dangerous failure is not losing position but not knowing you have.
The dangerous failure is a confident one
Satellite navigation is the easiest thing on a vehicle to take away. A jammer costs less than the receiver it defeats, and spoofing is no longer exotic. Neither is required: an urban canyon, a tunnel portal, a steel structure or a hangar roof will do it.
A vehicle that knows its position is bad slows, holds, or comes home. A vehicle that is wrong and certain flies into something. So we build the second number as carefully as the first — every fix carries a measured confidence, and the autopilot is wired to act on that confidence rather than on the coordinate alone.
Commercial receivers, integrated until they agree
We do not build GNSS silicon. We select it — multi-band, multi-constellation (GPS, Galileo, GLONASS, BeiDou) u-blox and equivalent modules — and then do the work that decides whether it performs: antenna placement, ground plane, interference rejection, and the code that reconciles it with every other sensor on the vehicle.
- Anti-jam and anti-spoof GNSS integrated into the airframe rather than bolted to it, with integrity monitoring that refuses a compromised fix instead of using it.
- Multi-sensor fusion across inertial, visual, barometric, magnetic, ranging and odometric sources, each weighted by measured confidence rather than by a constant somebody chose once.
- GNSS-denied (GPS-denied) navigation at centimetre-class relative accuracy against surveyed markers and visual references, for work where satellite positioning is unavailable, unreliable or untrusted.
- RTK where the application earns the infrastructure — and a written, tested answer for the moment the correction stream stops.
We test the transitions, not just the states
A fusion stack tuned on an open airfield will disappoint indoors, under canopy, near steel and in the cold. We characterise the environment the vehicle will actually work in, and tune against that one.
Then we fly the edges. Entering and leaving a degraded sector is where most implementations misbehave: the fix returns, disagrees with the inertial solution by half a metre, and the filter takes the jump. On our vehicles that transition is a test case with a pass criterion, not something discovered in service.
What you get
- Accuracy you can quote with a bound, in the environment you fly in
- A bad fix refused rather than flown on
- A written, tested behaviour for the moment positioning goes
Related work
Tell us the constraint you cannot get past
The mass budget, the latency, the environment, the certification route. We will tell you honestly whether it is something we should be working on — and if it is not, who should.