A quantum sensor flew a plane for four hours with GPS off — 89% better than the backup

An Embraer 170 flew from Puget Sound to southern Alaska and back — 4 hours 23 minutes over the Pacific — with GPS unavailable, navigating on Honeywell's MagNav system. MagNav uses quantum magnetometers to read the magnetic signature of the Earth's crust and match it against a map, giving a continuous position reference that does not depend on weather, time of day or a satellite signal. The Defense Innovation Unit reports position accuracy improved by 89% against the conventional backup methods carried on the same flight. Navigation data streamed to the pilots on standard tablets, so deploying it does not require rebuilding a cockpit. The programme began in spring 2024 as "Transition of Quantum Sensing"; 72 companies applied and Honeywell Aerospace was selected to build the prototype. A demonstration aboard a US Air Force C-17 is planned later this year. The tracker records the flight.

Why it matters

Quantum computing gets the coverage and quantum sensing gets the deployments. The distinction matters because they have opposite maturity curves: a quantum computer has to beat classical hardware at a useful task before it is worth anything, and that argument is still being fought contract by contract. A quantum magnetometer only has to be more accurate than an inertial unit drifting without GPS, which is a low bar cleared by physics that already works at room temperature and altitude.

The 89% figure is the useful number precisely because it is comparative. It is not "our sensor is accurate", it is "against the backup you actually fly today, over four hours, this was 89% better" — measured by the customer, not the vendor, on a real airframe over open ocean. That is the kind of evidence the quantum-computing side of the field has struggled to produce, and it arrived without a benchmark dispute attached.

The strategic reading is straightforward. GPS is jammable and spoofable, and the assumption that it is always there is baked into modern aviation. A navigation reference that reads the planet itself cannot be jammed, because there is no signal to attack. The caveat is coverage: MagNav needs a magnetic map of what it is flying over, and the quality of that map is the real constraint on where this works.

What to watch

Whether the C-17 demonstration holds the 89% margin on a larger, noisier airframe, how much of the world has crustal magnetic maps good enough to navigate on, and whether any civil operator picks it up.