Accuracy is addictive


Mar 11, 2001
Reaction score
Accuracy is addictive

Mar 14th 2002

From The Economist print edition

The invention of GPS married ideas from quantum mechanics and relativity with the need to track Russian satellites. Most remarkable of all, the concept—now the basis of a $12 billion industry—was put together over a single weekend

EVER since humans took to exploring beyond their front yard, they have sought ways of determining where they are and where they are going. But until fairly recently, the best positioning tools available—landmarks, celestial guidance and dead reckoning—all had disadvantages. Some were too local or complicated; others too finicky or inaccurate. In her book “Longitude”, Dava Sobel wrote that an inability to establish their whereabouts reliably caused every great captain in the age of exploration, from Vasco da Gama to Sir Francis Drake, to become lost when at sea.

During the 1950s, with their first artificial satellite in the works, the generals of the former Soviet Union were mulling over the merits of building a satellite navigation system. The KGB agents dispatched to the United States concluded that the Americans had no interest in such a system (true at the time) and had the generals' plan dismissed as a bad idea. It took the launch of Sputnik in 1957 to shock America into action—and to start a desperate bid to catch up. In the process, they conceived a satellite navigation system called the global positioning system (GPS), which solved the problem, once and for all, of how to locate yourself anywhere on the globe in three dimensions.

It was during the Gulf war in 1991 that the world first began to hear about GPS in a big way. Using hand-held devices rushed into production by a small firms such as Magellan, coalition troops were able to find their way around the Iraqi desert with uncanny accuracy. Since then, GPS technology has moved into public life. Today, it is being applied with ever-increasing precision to such activities as exploring the heavens or navigating the maze of streets and freeways around Los Angeles in a rented car.

It is used to track fleets of vehicles on the highways and self-steering tractors in the fields; to keep weekend yachtsmen out of trouble at sea; and to check the health of dams and bridges around the world. But the best-kept secret about GPS is its value as a clock. Time-sensitive operations associated with banking, telecommunications and even power grids have come to depend on GPS. Already the basis of a $12 billion global industry, GPS is an example of a self-perpetuating innovation: the better it gets, the more uses people find for it.

How did it come into being? Within days of Sputnik's launch in October 1957, researchers around the world were able to determine the satellite's orbit by analysing variations in the frequency (ie, the Doppler shift) of the satellite's radio signal as it passed overhead. Frank McClure, then chairman of the Applied Physics Laboratory (APL) at Johns Hopkins University in Baltimore, saw the navigational possibilities of using the Doppler shift from a satellite to locate a receiver on earth.

The following year, with money from the Pentagon, scientists at APL began to develop a satellite navigation system called Transit. At the time, the nature of the ionosphere and the earth's gravity field—which can have a serious effect on radio signals—were still conundrums. But progress was swift. The United States Navy, which was looking for a way to launch ballistic missiles accurately from nuclear submarines, assumed responsibility for the system in 1958. Within six years, Transit was up and running.

Subsequently, other satellite navigation efforts were started within the American defence establishment. The Timation project at the Naval Research Laboratory was based on highly accurate clocks. Not to be left out, the United States Air Force was sponsoring a project known simply as 621B, which used a special signal to prevent jamming.

Analysing the Doppler shift of signals from Transit's half a dozen satellites allowed a submarine to determine its location to within 25 metres. But the system took up to quarter of an hour to get a fix, and it did not give continuous positioning. In 1973, the Department of Defence proposed consolidating all the satellite navigation projects into one that combined the best features of the existing systems. After some initial wrangling, the GPS concept was hammered out over a holiday weekend.

Today, GPS consists of 28 satellites orbiting 10,900 nautical miles above the earth. An upgrade programme continually adds more satellites to the constellation. A network of ground stations—in locations as far-flung as Hawaii and Ascension Island in the South Atlantic—tracks the satellites, updating their positions and time signals to ensure their accuracy.

In essence, GPS works by measuring the time it takes a radio signal from a satellite to reach a receiver on the ground. Each satellite continuously broadcasts a signal that gives its position and the time. A GPS receiver compares its own time with the satellite's time, and uses the difference between the two to calculate the distance. Taking measurements from four satellites allows the receiver to pinpoint latitude, longitude and altitude, and to correct for errors in its clock, which is not nearly so precise (or costly) as the clocks in the satellites. The problem is simply one of solid geometry: finding the point of intersection of four spheres, each centred on one of the satellites. Using four signals is better than three; and even more is better still. Modern GPS receivers pluck down as many signals as possible.

Success, of course, depends on knowing exactly where the satellites are at any given moment. That part is not all that hard, since the satellites are constantly updated with navigational data by the control stations on the ground. But accurate timing is also critical for GPS. It takes less than a tenth of a second for the signal from a satellite overhead to reach a receiver on the ground. What device could keep time that precisely? The needed innovation, oddly enough, emerged from basic research of the most arcane sort.

Not your father's watch

If you have something that vibrates with a known frequency, you have the basis of a clock. Pendulums, springs, tuning forks and piezoelectric crystals have all been pressed into service with varying degrees of accuracy. But none compares with the oscillations that occur deep within the atom. The theory of quantum mechanics says that atoms will emit electro-magnetic radiation of a very specific frequency (“resonant frequency”) depending on the particular energy states that are available to that individual atom.

In 1944, the year he won a Nobel prize for “magnetic resonance”, Isador Isaac Rabi of Columbia University suggested that an atomic clock could be made using the resonant frequencies of caesium atoms. Rabi was interested in the properties of atomic nuclei, not clocks, and he never pursued the idea. But after the second world war, his former student, Norman Ramsey, improved Rabi's method, collecting his own Nobel prize for the work in 1989.

Creating atomic-time standards became a priority in both America and Britain after the second world war. Jerrold Zacharias at the Massachusetts Institute of Technology developed a practical atomic clock around the same time that Louis Essen and John Parry established the first atomic clock at the National Physical Laboratory in Teddington, west of London. Thanks to this pioneering work, “NIST-F1”, America's primary time and frequency standard at the National Institute of Standards and Technology in Boulder, Colorado, is accurate to one second in 20m years.

Each GPS satellite carries four atomic clocks synchronised to a universal standard. Each satellite transmits information at two frequencies, L1 and L2. The signals include “pseudo-random codes”, which are a sequence of “up” and “down” states that look like random electrical noise. The innovation of assigning a unique pseudo-random code to each satellite allows all the satellites in the GPS constellation to use the same frequencies without jamming one another, and protects them from other interfering signals as well. As a bonus, the random code's continuous signal provides precise information about the time.

Remarkably, GPS calls into play Einstein's theories of special and general relativity—among the most esoteric theories of physics. Because of the velocity with which the satellites orbit the earth and the weaker gravitational field that they move through, the atomic clocks on board the GPS satellites advance faster than comparable clocks on the ground. If the effects of relativity were not taken into account, GPS would not work.

It was always assumed that the public would get access to GPS one day. But it was not until the early 1990s that a sufficient complement of satellites was in orbit. While the military signals were accurate to 18 metres and encrypted for exclusive use, the civilian signals were distorted deliberately to reduce their accuracy to 100 metres. President Clinton ended the restrictions on accuracy for commercial use in May 2000. Today, the public can get GPS position fixes with an accuracy of three metres to 15 metres, depending on where they are.

Actually, GPS can do a lot better than that. By cross-referencing the satellite data with a ground-based transmitter of precisely known location, the accuracy can be boosted tenfold or more. Aviation authorities around the world have been testing “differential GPS” for blind landing in fog, where approaching aircraft need to be positioned vertically and horizontally with an accuracy of ten centimetres or less. Then there is the “advanced GPS” that is used for scientific and other precision measurements such as monitoring seismic activity and ocean currents, and for remote sensing of global warming within the atmosphere. Using additional triangulation, this can pinpoint an object's position to within one centimetre.

Thanks to Sputnik

At a time when American rockets were still toppling over on the launch-pad, the idea of using satellites to navigate was as audacious as it was ingenious. To start with none, and then to send a whole constellation of satellites into space within a few decades, represents a remarkable feat of technology. Establishing the vision and the means for organising, tracking and maintaining a project of such global proportions and complexity was equally impressive.

GPS is a classic example of the process of technical evolution. Observation of Sputnik's signals led to Transit, the world's first satellite navigation system. Transit sparked competing efforts to develop an even better navigational system. In turn, that lead the Pentagon to consolidate all the projects into one focused international (though American-funded) push. And out of that came GPS.

As a case history of innovation, GPS shows the importance of being able to understand the physics at the most theoretical level while also being able to envision the technology needed to use it practically, and then ruthlessly prevent roadblocks, rivalries or lack of focus to hinder progress. It is doubtful that the global positioning industry would be where it is today without the Pentagon's role as midwife. Nevertheless, the lessons learned from GPS's commercialisation apply equally to technological innovation in the biggest and smallest of companies.

Top Bottom