Skip to content

All news

Insight

A position is only as good as its error bound

Magnetic navigation has already flown. Several teams have shown that reading the Earth's magnetic field can correct a vehicle's position without satellites. The question is no longer whether it works.

The open question is trust. A position on its own leaves the pilot, or the autopilot, to guess how far to rely on it. What matters is not how accurate the system was on a good day. It is how wrong the position can be right now.

Civil aviation already works this way with satellite navigation. For an instrument approach, the receiver computes a protection level: a bound that its position error should exceed only with a very small, specified probability. The approach is available only when that bound fits the operation. Magnetic navigation needs the same discipline.

We design our instrument around that idea. Every fix comes with its error bound, computed at every instant. Integrity: the instrument says when not to trust it.

It completes the inertial unit, it does not replace it. Between fixes, the inertial unit carries the position; each magnetic fix corrects it and comes with its bound.

The bound is declared by the instrument itself. Certifying it is the work of a programme, with its users and its authority. Until our first prototype is assembled and tested, what we show is model-derived, from simulation.

Insight

Satellite navigation is contested every day

Jamming and spoofing of satellite positioning are part of everyday flight operations in several regions. The figures, with their sources, and what they mean for navigation.

Insight

Rejecting the vehicle's own field, on board

Magnetometers flying today are corrected by compensation models fitted to each vehicle. We design our sensor to reject the platform's own field on board as well.