GNSS is superb outdoors and useless in a tunnel — and the position still has to be right. Holding an accurate location through urban canyons, underpasses and indoor stretches means fusing satellite fixes with inertial and motion sensors so that when the sky disappears, the estimate coasts instead of jumping.
Satellite positioning is excellent with a clear sky and degrades sharply without one. In a city the signal reflects off buildings and the fix jumps; in a tunnel or a car park it vanishes entirely. For navigation, tolling, tracking or an autonomous function, a position that disappears or leaps is not acceptable.
The answer is not a better antenna but sensor fusion. An inertial measurement unit, wheel or motion data and a model of how the platform moves let the system dead-reckon through the gap, while the satellite fix, when trustworthy, corrects the drift. The art is trusting each source exactly as much as it deserves, moment to moment.
So the system is built as a fusion problem: honest sensor models, a filter that weights satellite and inertial data by their real reliability, and detection of the reflected, untrustworthy fixes that would otherwise poison the estimate.
Stated before any of them had an answer.
Tunnels, canyons and indoors are where the fix vanishes or reflects; the system has to be best precisely where the satellites are worst.
In a city, multipath produces a satellite fix that looks valid and is metres wrong; trusting it naively is worse than ignoring it.
Dead reckoning coasts through gaps but drifts over time; it holds position for a while, not forever, and the fusion has to know that.
The right weighting between satellite and inertial changes second to second with conditions; a fixed blend fails in the hard moments.
A clear-sky fix is easy. The engineering is the tunnel, the canyon and the car park — coasting on inertial and rejecting the reflections until the satellites come back.
The scope of the work, rather than a capability list.
Rarely the subsystem that sounds difficult.
The core difficulty is deciding, second by second, how much to believe the satellite versus the inertial data as conditions change; a fixed answer fails.
A multipath fix in a city looks entirely valid; recognising and rejecting it before it corrupts the estimate is subtle and essential.
Dead reckoning is only as good as how long it holds; understanding and limiting the drift is what makes coasting through a tunnel trustworthy.
When the sky returns the estimate has to snap back to truth without a visible jump, which the fusion has to manage deliberately.
A continuous location through tunnels, canyons and indoors, coasting where the satellites fail.
Reflected, untrustworthy fixes rejected instead of poisoning the estimate.
Dead reckoning that holds accurately for the length of a real gap.
A position that corrects smoothly when GNSS returns rather than jumping.
Customer projects are presented at property, capability, outcome and integration level. Customer names, internal architecture, confidential deliverables and commercial terms are not disclosed. Where a detail would identify a customer it is omitted rather than approximated. More is available under a non-disclosure agreement, within the limits the customer has agreed.
Every item links to its own page.
Because the problem is not sensitivity but physics. In a tunnel there is no satellite signal to receive, however good the antenna, and in an urban canyon the signal reflects off buildings so the fix is present but wrong. No receiver improvement fixes a signal that is absent or corrupted, which is why the answer is fusing GNSS with inertial and motion sensors that keep working when the sky does not.
Dead reckoning estimates position from motion — how fast and in what direction the platform has moved since the last known fix — using inertial and vehicle sensors. It lets the system coast through a GNSS gap, but it drifts: small sensor errors accumulate over time, so it holds an accurate position for the length of a tunnel or an underpass, not indefinitely. The fusion has to understand and bound that drift and correct it the moment a trustworthy satellite fix returns.
By weighting each source by its real reliability and actively detecting multipath. In a city, a reflected signal can produce a fix that looks valid but is metres off, and blindly trusting it is worse than ignoring it. The fusion filter recognises the signatures of an untrustworthy fix and down-weights or discards it, leaning on the inertial estimate until a genuinely reliable fix is available.
Written for engineers. Share it with one.
Tell us the specification, the constraint and the deadline. Programmes that cross silicon, radio, embedded and AI are where Faststream is strongest.