How a GPS Speedometer Works on iPhone (and How Accurate It Is)

Your car measures speed by counting how fast a wheel turns. Your iPhone does something completely different: it listens to satellites, works out where it is several times a second, and derives how fast that position is changing. The two methods disagree in interesting ways, and knowing why makes the number on your screen far more useful.

Position, then speed

A GNSS receiver — GPS in the United States, plus Galileo, GLONASS and BeiDou, all of which modern iPhones use together — listens for timing signals broadcast by satellites. Each signal carries the exact moment it was sent. By comparing that against its own clock, the receiver works out how far away each satellite is. With four or more distances it can solve for latitude, longitude, altitude and a time correction.

Do that once a second and you have a track. Speed is the distance between consecutive fixes divided by the time between them. In practice the receiver also uses the Doppler shift of the incoming signals — the same effect that changes the pitch of a passing siren — which gives a smoother and more accurate velocity than differencing positions alone. This is why a good GPS speed reading is often steadier than you would expect from the position accuracy.

What actually limits accuracy

Three things dominate.

Satellite geometry. If the satellites you can see are clustered in one part of the sky, the mathematics is poorly conditioned and small timing errors turn into large position errors. Spread across the sky, the same errors mostly cancel. Receivers report this as dilution of precision.

Multipath. Signals bounce. In a street lined with glass towers your receiver may be hearing a reflection that travelled an extra hundred metres, and it has no easy way to know. Multipath is the main reason city-centre readings wander.

Obstruction. Concrete, earth and dense foliage stop the signal outright. In a tunnel or an underground car park there is nothing to receive, so the reading freezes or drops out entirely until you emerge.

Getting a reading you can trust

• Give it a clear view. A phone mounted on the windscreen or dashboard sees far more sky than one lying in a cup holder or a pocket.
• Wait for the fix to settle. The first few seconds after launch are the least reliable. Let the number stabilise before you judge it.
• Judge it at steady speed. Comparing readings while accelerating hard is comparing two instruments with different response times, which tells you nothing about either.
• Expect tunnels to break it. This is physics, not a fault. The reading will recover within seconds of regaining sky view.

Reading the numbers in the app

Speedometer shows the live figure three ways — as a plain digital readout, as an analog dial with a scale you can set up to 0–210, and as a mirrored HUD you can reflect off the windscreen at night. The number is identical in all three; only how quickly you can read it changes.

While a trip is running the app also tracks distance, average speed and maximum speed for that session, and when you stop, the trip is filed in History with its route drawn on a map and the start and finish locations recorded. That record is more forgiving than the live number: a one-second glitch barely moves an average computed over twenty minutes.

The honest summary

A phone GPS is an excellent speedometer for the things drivers actually care about — is my speed drifting up, how fast was that stretch, how far did I go — and a poor one for anything demanding certified accuracy. It measures your true motion over the ground rather than wheel rotation, it needs a view of the sky, and it is a driver aid rather than an instrument of record.