Your phone’s blue dot looks like a map trick. Underneath it is a timing machine: a swarm of satellites carrying atomic clocks, radio signals that travel at the speed of light, and a little computer that turns tiny time differences into a place on Earth.
That system is the Global Positioning System, or GPS. It began as a U.S. military program called NAVSTAR. Today it is an open utility that ships, planes, phones, and mapping apps lean on every day — not because someone “tracks” you from space, but because distance can be measured as time.

Think of it this way. Light moves about 300,000 kilometers per second. A billionth of a second of timing error is roughly 30 centimeters of distance error. So GPS does not mainly “see” you. It times you.
What follows is how that machine works: satellites and clocks, trilateration in plain language, why Einstein’s relativity shows up in the engineering, and how a Cold War navigation system became civilian infrastructure after Selective Availability ended in 2000.
What is flying overhead
GPS is three cooperating pieces. The space segment is the satellites. The control segment is ground stations that watch the clocks and orbits and upload corrections. The user segment is everything that listens: phones, car units, ship bridges, survey gear, aircraft receivers.
The U.S. Space Force flies the constellation. Official policy is to keep at least 24 operational satellites available 95% of the time. In practice the force has flown about 31 operational birds for years, with extras covering maintenance and aging craft. They orbit at roughly 20,200 kilometers — medium Earth orbit — and each one circles the planet about twice a day.
The geometry is deliberate. Satellites sit in six orbital planes. The classic “24-slot” layout puts four slots in each plane so that, almost anywhere on Earth, you can see at least four satellites. Four is the magic number for a full three-dimensional fix, for a reason we will get to in a moment.
Each satellite carries redundant atomic clocks — ultra-stable timekeepers that use the natural vibration of atoms (commonly rubidium or cesium standards in GPS hardware). Those clocks are the heart of the system. The satellite is constantly broadcasting: “Here is my time, and here is where I am in my orbit.” Your receiver’s job is to compare those messages.
From a time stamp to a distance
A GPS signal is a crafted radio broadcast. Encoded in it is a precise time mark and a navigation message that describes the satellite’s orbit (its ephemeris) and clock corrections. Your phone does not need its own atomic clock. It needs a decent quartz oscillator and enough math.
Here is the everyday idea. The satellite sends a time stamp. Your receiver notes when that stamp arrives. Multiply the travel time by the speed of light, and you get a distance — the range to that satellite.
In practice that measured distance is called a pseudorange, because your receiver’s cheap clock is slightly wrong. A tiny clock bias turns into a huge distance bias if you are not careful. The elegant fix is to treat the receiver’s clock error as a fourth unknown and solve for it along with latitude, longitude, and altitude.
That is why four satellites are the usual minimum. Three good ranges would locate you in space if your clock were perfect. The fourth range lets the receiver slide its own clock forward or back until the spheres of distance agree on one point. More satellites make the answer stabler when some signals are noisy or blocked.
Trilateration without the jargon fog
People often say “triangulation.” GPS is closer to trilateration: finding a point from distances, not from angles.
Picture one satellite. Knowing you are 20,000-plus kilometers away puts you somewhere on a huge sphere centered on that satellite. A second satellite gives a second sphere. Two spheres intersect in a circle. A third sphere cuts that circle down to (usually) two points — one of which is nonsense (far out in space) and one of which is near Earth’s surface. A fourth measurement settles the clock error and tightens the answer.
Your phone then paints that coordinate onto a map made by someone else. GPS itself is the positioning and timing service. The streets, traffic layers, and turn-by-turn voice are applications riding on top.
Accuracy is not one number. Under open sky, a typical smartphone is often within about 4.9 meters (roughly 16 feet), according to GPS.gov’s civilian accuracy summary. Canyons of buildings, heavy tree cover, and indoor use make it worse. Surveyors and aviation systems add dual-frequency receivers and augmentation services to reach centimeters when they need them.

Why relativity is not optional
GPS satellites move fast and sit higher in Earth’s gravity well than you do. Einstein’s theories say both facts change how clocks tick.
Special relativity says a moving clock runs slow relative to a clock at rest. For GPS orbits, that effect alone would make the satellite clocks fall behind ground clocks by about 7 microseconds per day.
General relativity says clocks run faster where gravity is weaker. Higher up, GPS clocks would gain about 45 microseconds per day compared with clocks on Earth.
Net result: without correction, satellite clocks run about 38 microseconds per day fast relative to ground time. A microsecond is a millionth of a second. At light speed, 38 microseconds is more than 10 kilometers of ranging error per day if you ignored it. The system would drift into uselessness.
So engineers pre-adjust the satellite clock rates — a fixed fractional frequency offset on the order of −4.46×10⁻¹⁰ for the circular-orbit case, in the NIST technical literature — and receivers also apply a small eccentricity correction when orbits are not perfectly circular. Relativity is not a trivia sidebar. It is baked into the interface control documents the receivers follow.
NIST has called GPS one of the longest-running practical tests of Einstein’s predictions: the blue-dot economy works because the corrections work.
From NAVSTAR to an open civilian utility
GPS grew out of earlier U.S. military satellite navigation experiments and was developed as NAVSTAR GPS. The design goal was global, all-weather positioning for armed forces — ships, aircraft, ground units — with precise timing as a twin product.
Civilian receivers could listen to a public signal, but for years the U.S. applied Selective Availability (SA): an intentional degradation of that civil accuracy for national-security reasons. In the 1990s, everyday GPS was useful but fuzzy compared with what the hardware could do.
On May 1, 2000, at President Bill Clinton’s direction, the government turned SA off for the whole constellation at once. GPS.gov’s record is blunt: the United States has no intent to use Selective Availability again, and later GPS III satellites were procured without the SA feature so the policy would stick in hardware.
That single engineering-and-policy choice mattered as much as any new satellite. Overnight, civil receivers worldwide jumped in accuracy without a hardware swap. Shipping companies, aviation planners, farmers, surveyors, and phone makers could build products on a freer signal. Differential and augmentation systems still matter for safety-critical and centimeter work, but the baseline civil service stopped being artificially dulled.
Why the machine remade shipping, flying, phones, and maps
Once position and time are cheap, whole industries reorganize around them.
At sea, GPS replaced or backed up older celestial and radio methods with continuous fixes in fog and dark. On the flight deck, it feeds navigation and approach procedures (often blended with other sensors and integrity monitors). On land, trucking, rail, and last-mile delivery treat location as inventory.
Phones made the user interface universal. The same ranging math that guided a destroyer now underpins rideshare pins, hiking tracks, and “share my location.” Mapping firms fuse GPS traces with imagery and road databases. Emergency services use handset location as a starting guess. Financial networks and cell towers also lean on GPS timing, not just blue dots — synchronized clocks keep packets and trades orderly.
None of that requires the satellite to know who you are. The broadcast is one-way. Your device listens, computes, and (if an app chooses) may send the result somewhere else. The engineering breakthrough was making precise time a public utility in orbit.
What still limits the blue dot
GPS is astonishing and imperfect. Skyscrapers bounce signals (multipath). Indoor use is weak. Solar storms and interference can degrade ranging. The government commits to signal-in-space performance standards; your actual meter-level error still depends on geometry, atmosphere, and the chip in your hand.
Other constellations — Europe’s Galileo, Russia’s GLONASS, China’s BeiDou — now share the sky, and modern phones often use several systems together under the broader label GNSS (global navigation satellite system). The core idea stays the same: flying clocks, light-speed ranging, and a receiver that solves for where it must be.
So when the map snaps to your corner, you are not being “found” by a camera in space. You are timing echoes from atomic clocks in medium Earth orbit, correcting for relativity, and turning four or more distances into a point. That is the machine. The blue dot is just the friendly face on top.
Further reading

Pinpoint: How GPS is Changing Technology, Culture, and Our Minds — Greg Milner’s accessible history of GPS — from Cold War guidance to the blue dot on every phone — and what always-on location does to culture and attention.

You Are Here: From the Compass to GPS, the History and Future of How We Find Ourselves — A wider tour of how humans find their way, ending in satellite navigation and the privacy questions that follow.

Global Positioning System: Theory & Applications (Volume One) — The classic Parkinson/Spilker technical reference for readers who want the engineering depth behind ranging, clocks, and receiver design.