Skip to content
Void Lattice / Labs / Quantum Navigation

Quantum Navigation

The crust is magnetised unevenly. Buried rock, old lava flows and ore bodies leave a pattern in the field a few tens of nanotesla deep, fixed in place and different everywhere — and it has been surveyed from the air over much of the world. A drone that has lost GPS can still measure it, and a sequence of those measurements, matched against the survey, can say where the drone is.

Nitrogen-vacancy diamond magnetometers read that field through the resonance of a single electron spin, with a lower noise floor and a scale fixed by an atomic constant. The lab flies both kinds of sensor through the same simulated world, with the same inertial unit, map and weather, and lets the navigation filter show when the quantum sensor's precision becomes position accuracy, and when the map, the airframe or the ground itself decide the answer instead.

Ghost Lattice takes it to a group: one beacon drone carries the quantum magnetometer, cheap followers carry radios and range to it and to each other, and the group keeps one shared reference through a GPS-denied canyon — with rock that blocks radio paths, a rogue drone, a jammer that cuts every link, and a saddle of rock it has to climb over.

Where it holds sets the quantum magnetometer beside LiDAR-inertial odometry, the GPS-free navigation of the forest drones: one flight across forest, open water, a magnetically quiet basin, smoke and a high transit, four navigators on the same sensor records, reviewed on strip charts with a playhead you can drag.

01

The simulator

What if a drone could navigate using Earth's invisible magnetic fingerprints?

VL-X04 · Field → sensor → map matching → position

Quantum Navigation

Disable GPS. Measure the magnetic field. See whether quantum sensing can find the way home. Five experiments — a GPS blackout, quantum against classical, magnetic ghosts, a featureless Earth, and a drone dropped somewhere unknown — on a live navigation filter.

Open Quantum Navigation
VL-X04 · One beacon → radio ranging → a shared reference

Ghost Lattice

Lost on Earth, found by each other. A beacon with the quantum magnetometer leads cheap followers through a GPS-denied canyon by radio ranging. Blind followers, rock that blocks the radio, a rogue drone, a jammer that splits the group, a ridge to climb over, and a head-to-head against a conventional GPS and radio fleet — in 2D or 3D.

Open Ghost Lattice
VL-X04 · LiDAR odometry ↔ the field → where each one holds

Where it holds

LiDAR or the field? A LiDAR-inertial odometer and an NV magnetometer, alone and together, on one flight across forest, open water, a magnetically quiet basin, smoke and a climb out of LiDAR range. A flight review with strip charts and a playhead, and eight seeds to check it.

Open Where it holds
02

What it shows

THE SIGNAL

A fingerprint a few tens of nanotesla deep

Earth's field is about 50 000 nT. Buried magnetised rock bends it by tens to hundreds of nanotesla, in a pattern that is fixed in place and different from one kilometre to the next. Fly higher and the pattern fades, its finest detail first. That pattern, surveyed from the air, is the map a drone without GPS can navigate against.

THE SENSOR

A spin that moves 28 Hz per nanotesla

A nitrogen-vacancy centre in diamond has a spin whose resonance sits near 2.87 GHz and shifts with the field. Green light makes it glow red; a microwave on resonance dims the glow. Reading where the dips fall is a magnetometer with a low noise floor and a scale set by an atomic constant. It measures the field, and only the field.

THE MATCH

Many places at once, until only one fits

A reading of 50 214 nT is consistent with every place the map reads 50 214. The navigator is a marginalised particle filter: thousands of position hypotheses, each carried forward by the inertial unit and weighed against the map, so several separate candidates can survive until the drone crosses ground distinctive enough to choose between them.

THE LIMITS

Sensitivity is not accuracy

Position comes from how the field changes with position, not from how precisely it is read. A perfect sensor over flat ground, or reading a poor map, learns nothing; a magnetically noisy airframe or a hot sensor bay can undo the quantum advantage entirely. The lab is built so that each of those can be seen to happen.

Educational simulation. The magnetic world, the survey map and both sensors are modelled, and the sensor figures are illustrative rather than taken from any device; the navigation filter is a real estimator running on those simulated measurements. No result here is a measured or published performance figure, and the NV magnetometer is not presented as a replacement for GPS.