A Small Mistake, A Big Problem
On September 23, 1999, NASA lost contact with the Mars Climate Orbiter, a spacecraft that had traveled 416 million miles over nine and a half months to reach the Red Planet. As the orbiter fired its engines to slip into orbit around Mars, the flight controllers at the Jet Propulsion Laboratory in Pasadena watched their screens with growing unease. The spacecraft was descending far too low. Within minutes, it plunged into the Martian atmosphere at an altitude of just 57 kilometers, when it should have passed safely overhead at 150 kilometers. The orbiter was torn apart and destroyed.
The subsequent investigation revealed a cause so mundane it was almost embarrassing. Lockheed Martin, the company that had built the spacecraft, had written its thruster software using imperial units, specifically pound-force seconds. NASA's navigation team at JPL, meanwhile, had been expecting the data in metric units, specifically newton seconds. Nobody had caught the discrepancy during the months of flight. Every tiny course correction had been slightly wrong, and those errors had accumulated over millions of miles until the spacecraft arrived at Mars on a trajectory that was fatally off.
The mission had cost 327.6 million dollars. The root cause was a failure to convert between pounds and newtons.
This story is well known in engineering circles, where it has become the textbook example of why unit consistency matters. But the Mars Climate Orbiter is far from the only time a measurement mix-up has had dramatic consequences. History is littered with disasters, near-disasters, and strange twists of fate that all trace back to someone getting their units wrong. Some of these stories are terrifying. Some are surprisingly funny. And a few turned out better than anyone had a right to expect.
The Gimli Glider: Running on Empty at 41,000 Feet
On July 23, 1983, Air Canada Flight 143 departed Montreal for Edmonton with 69 people on board. Partway through the flight, at an altitude of 41,000 feet over the vast emptiness of the Canadian prairies, both engines went silent. The Boeing 767, one of the most advanced commercial aircraft in the world at the time, had run out of fuel.
The chain of events that led to this moment was a comedy of errors rooted in Canada's then-recent switch from imperial to metric units. The 767 was one of Air Canada's first metric aircraft, with fuel gauges calibrated in kilograms rather than pounds. On the ground in Montreal, the fuel quantity indicator system had malfunctioned, so the ground crew needed to calculate the fuel load manually. They knew the plane needed 22,300 kilograms of fuel for the flight. They dip-tested the tanks and found that 7,682 liters of fuel were already on board. The question was simple: how many more liters did they need to add?
But to answer that question, they had to convert between liters and kilograms, and this is where everything fell apart. The crew used a conversion factor of 1.77, which was the number they had always used. The problem was that 1.77 was the conversion factor for pounds per liter, not kilograms per liter. The correct factor for kerosene jet fuel in kilograms per liter is 0.803. By using the wrong number, the crew calculated that far more fuel was already on board than actually was. They added only a fraction of what the plane needed.
What followed was one of the most remarkable feats of piloting in aviation history. Captain Robert Pearson, who happened to be an experienced glider pilot, realized he could reach a decommissioned Royal Canadian Air Force base at a place called Gimli, Manitoba. He guided the powerless 767 through a dead-stick approach and landed it on a runway that had been converted into a drag racing strip. A community barbecue was taking place on the old taxiway. Miraculously, nobody was seriously injured, though the nosewheel collapsed because the landing gear, without hydraulic power, could not fully lock into place.
The aircraft was repaired, returned to service, and continued flying for another 25 years. It was finally retired in 2008, and to this day it is remembered as the Gimli Glider. The incident led to sweeping changes in how airlines handle fuel calculations during the transition between measurement systems, and it remains one of the most vivid illustrations of how a simple unit conversion error can cascade into a life-threatening emergency.
Columbus and the Small Earth
Long before spacecraft and jet fuel, one of history's most consequential measurement errors set in motion the European colonization of the Americas. When Christopher Columbus sailed west from Spain in 1492, he was not operating on blind faith. He had done calculations. The problem was that his calculations were spectacularly wrong, and the error came down to mixing up two different kinds of miles.
Columbus relied on the work of earlier geographers, particularly the 9th-century Persian astronomer Alfraganus, who had estimated the Earth's circumference at roughly 20,400 miles. This estimate was actually quite accurate for the time. The trouble was that Alfraganus had been using Arabic miles, which were significantly longer than the Roman miles that Columbus was thinking in. An Arabic mile was about 1,975 meters, while a Roman mile was only about 1,480 meters. Columbus never accounted for this difference.
The result was that Columbus believed the Earth was about 25 percent smaller than it actually is. He calculated that the distance from the Canary Islands to Japan was roughly 3,700 kilometers. The real distance is closer to 19,600 kilometers, more than five times what he expected. Had the Americas not happened to be in the way, Columbus and his crew would have sailed into open ocean until their supplies ran out, and the voyage would have ended in death rather than discovery.
In a sense, the European encounter with the New World was made possible by a unit conversion error. Columbus was wrong about almost everything, but he was lucky, and sometimes in history that is enough.
The Vasa: A Warship Sunk by Two Rulers
In 1628, the Swedish warship Vasa set sail on its maiden voyage in Stockholm harbor. It was the pride of the Swedish navy, one of the most heavily armed vessels in the world, built over two years at enormous expense. It sailed for about 1,300 meters before a gust of wind caught its sails, and it heeled over and sank in full view of a horrified crowd of spectators. Thirty people drowned.
When the Vasa was salvaged from the bottom of Stockholm harbor in 1961, remarkably well preserved by the cold, brackish water of the Baltic, archaeologists were finally able to investigate what had gone wrong. They made a surprising discovery: the ship was measurably asymmetric. The port side was heavier than the starboard side. The hull planking was thicker on one side than the other. The ballast had been loaded unevenly in an attempt to compensate, but it was not enough.
The most likely explanation traces back to the fact that two teams of shipwrights had worked on opposite sides of the vessel. Archaeological evidence shows that the workers on one side were using Swedish feet (about 29.69 centimeters) while the workers on the other side were using Amsterdam feet (about 28.31 centimeters). The difference of roughly 1.4 centimeters per foot might sound trivial, but applied consistently across the entire length, width, and depth of a large warship, it was enough to make the vessel dangerously lopsided. Combined with the ship's already top-heavy design (the king had insisted on an extra gun deck), the asymmetry was fatal.
The Vasa sat on the seabed for 333 years. Today it is displayed in its own museum in Stockholm, where it stands as one of the world's most beautifully preserved examples of what happens when two groups of people measure things differently and nobody checks whether their numbers agree.
The Kilogram That Lost Weight
For more than a century, the kilogram was defined by a single physical object: a small cylinder of platinum-iridium alloy, about the size of a plum, stored under three nested bell jars in a climate-controlled vault at the International Bureau of Weights and Measures just outside Paris. This object, known as the International Prototype of the Kilogram or simply Le Grand K, was the kilogram. By definition, it weighed exactly one kilogram, and every other weight measurement in the world was ultimately calibrated against it.
The problem was that Le Grand K was changing. When scientists compared it against its official copies (stored in national metrology labs around the world), they found that the copies were gradually diverging from the original. Over the course of a century, the difference had grown to about 50 micrograms, roughly the weight of a fingerprint. It was impossible to tell whether Le Grand K was getting lighter or the copies were getting heavier, because there was no independent reference to check against. The kilogram was defined as whatever Le Grand K happened to weigh at any given moment, which meant that if atoms were slowly escaping from its surface, the entire world's system of mass measurement was drifting along with it.
This situation troubled metrologists for decades. The meter had long since been redefined in terms of the speed of light, and the second was defined by the vibrations of cesium atoms, but the kilogram remained stubbornly tied to a physical artifact that was, quite literally, losing weight.
In 2019, the problem was finally solved. The kilogram was redefined in terms of the Planck constant, a fundamental value in quantum mechanics that never changes. The new definition ties the kilogram to the laws of physics rather than to a lump of metal in a Parisian suburb, and it means that a perfectly accurate kilogram can now be realized anywhere in the universe without reference to any particular object. Le Grand K is still in its vault, still carefully maintained, but it is no longer the kilogram. It is just a very well-made piece of metal that weighs very close to one kilogram.
The Hubble Telescope's Blurry Vision
When the Hubble Space Telescope was launched in April 1990, it was supposed to revolutionize astronomy by providing images of unprecedented clarity from its vantage point above the distorting effects of Earth's atmosphere. Instead, the first images it sent back were blurry. Something was deeply wrong with the telescope's primary mirror.
The investigation revealed that the mirror had been ground to the wrong shape. It was exquisitely precise in its manufacture, polished to within a fraction of a wavelength of light, but it was precisely the wrong shape. The outer edge of the mirror was too flat by about 2.2 micrometers, roughly one fiftieth the thickness of a human hair. This tiny error, known as spherical aberration, was enough to render the telescope's images useless for many of the scientific observations it had been designed to perform.
The cause was a flaw in the testing equipment used during manufacture. A small measuring rod called a null corrector, used to check the mirror's curvature during polishing, had been assembled with one lens positioned 1.3 millimeters too far from where it should have been. The technicians at Perkin-Elmer, the company that built the mirror, had noticed anomalous results during testing but assumed their backup testing methods were less accurate than the null corrector. In fact, the backup methods were correct and the null corrector was wrong.
The error cost 1.5 billion dollars to fix. In 1993, astronauts on the Space Shuttle Endeavour installed a set of corrective optics, essentially giving the telescope a pair of glasses. The repair worked beautifully, and Hubble went on to become one of the most successful scientific instruments ever built, delivering decades of groundbreaking discoveries. But it very nearly became the most expensive piece of orbiting junk in history, all because of a measurement error smaller than the width of a pin.
The Metric Mishap at 30,000 Feet
The Gimli Glider was not the last time aviation and metric conversion made for a dangerous combination. In 2004, a Korean Air cargo jet overran the runway in Halifax, Nova Scotia, after the flight crew miscalculated the required runway length. The crew had been given the runway length in feet but entered it into their flight computer in meters, which made the runway appear nearly three and a half times longer than it actually was. The aircraft crashed through a fence and came to rest in a wooded area. The crew survived, but the plane was destroyed.
These kinds of errors might seem difficult to make, but they are more common than most people realize. Aviation safety databases contain numerous incidents where altitude, distance, or fuel quantity was misunderstood because of unit confusion. The International Civil Aviation Organization mandates the use of specific units (feet for altitude, nautical miles for distance, knots for airspeed), but aircraft and airports around the world do not always follow these conventions consistently, and pilots flying international routes must constantly be aware of which system is in use at any given moment.
China and Russia, for example, historically used meters for altitude rather than feet, which created potential confusion at the boundaries between their airspace and neighboring countries that used feet. Transitions between measurement systems at altitude, where split-second decisions can mean the difference between safety and catastrophe, require constant vigilance and extremely clear communication.
A Pharmacy Error That Changed Medicine
Unit confusion in medicine is less dramatic than a spacecraft crashing or a plane running out of fuel, but it can be just as deadly. In 1999, the Institute of Medicine published a landmark report estimating that medical errors killed between 44,000 and 98,000 Americans every year, and a significant portion of those errors involved incorrect dosing of medications.
One recurring problem is the confusion between milligrams and micrograms, units that differ by a factor of 1,000. A dose written as 0.1 mg is the same as 100 mcg, but if a harried nurse or pharmacist misreads or miscalculates, the patient might receive a thousand times too much or too little of a critical drug. Similar risks exist with the old apothecary system, where drugs were occasionally prescribed in grains (one grain equals approximately 64.8 milligrams) alongside the metric system, creating opportunities for dangerous confusion.
These problems have driven a sustained push in healthcare to standardize on metric units, use clear and unambiguous abbreviations, and build automated checks into electronic prescribing systems. The effort has saved countless lives, but errors still occur, and they serve as a constant reminder that getting units right is not just an academic exercise. In medicine, a misplaced decimal point can be the difference between a cure and a catastrophe.
What the Meter Got Wrong About Itself
There is a delicious irony at the heart of the metric system's founding story. The meter was supposed to be one ten-millionth of the distance from the North Pole to the equator, a definition rooted in the geometry of the Earth itself. To establish this distance, the French government dispatched two astronomers, Jean-Baptiste Delambre and Pierre Mechain, on an epic surveying expedition that lasted seven years, from 1792 to 1799. They measured the arc of the meridian from Dunkirk in northern France to Barcelona in Spain, triangulating their way across the countryside during a period of revolution, war, and political upheaval. Mechain was briefly imprisoned in Spain. Delambre was repeatedly stopped by suspicious locals who thought his surveying instruments looked like weapons.
When they finished and calculated the distance, they derived a length for the meter that was cast into a platinum bar and deposited in the French National Archives. It was a triumph of Enlightenment science. There was just one problem: the Earth is not a perfect sphere. It is slightly flattened at the poles and bulging at the equator, and the degree of flattening is not uniform. Mechain actually discovered an inconsistency in his measurements that suggested the meridian was not quite the shape they had assumed, but he was so distressed by the discrepancy that he concealed it, adjusting some of his figures to make them fit the expected pattern. The deception was not discovered until after his death.
As a result, the original meter was slightly too short. The distance from the North Pole to the equator is not exactly 10,000,000 meters. It is closer to 10,001,966 meters. The error is tiny in relative terms (about 0.02 percent), but it means that the meter has never actually been what its creators said it was. In practice, this does not matter at all, because the meter was redefined long ago in terms of the speed of light, and its connection to the Earth's circumference is now purely historical. But there is something wonderfully human about the fact that the world's most rational system of measurement began with a surveyor who fudged his data because the real numbers made him anxious.
GPS: Where Nanoseconds Become Meters
Every time you open a map app on your phone, you are relying on one of the most precise measurement systems ever built, and its accuracy depends on getting units right at a scale that would have been unimaginable to the scientists who defined the meter two centuries ago.
The Global Positioning System works by measuring the time it takes for radio signals to travel from orbiting satellites to your phone. Since radio waves travel at the speed of light (roughly 300,000 kilometers per second), even a tiny error in timekeeping translates into a significant error in position. One nanosecond, one billionth of a second, corresponds to about 30 centimeters of distance. To achieve meter-level positioning accuracy, the atomic clocks on GPS satellites need to be synchronized to within a few nanoseconds.
But it gets more complicated. Einstein's theories of relativity predict that clocks in orbit tick at a different rate than clocks on the ground, because of differences in both speed (special relativity) and gravitational field strength (general relativity). The combined effect is that GPS satellite clocks gain about 38 microseconds per day relative to ground clocks. If this relativistic correction were not applied, GPS positions would drift by roughly 10 kilometers per day, rendering the entire system useless within hours.
The fact that your phone can pinpoint your location to within a few meters is a testament to humanity's ability to measure time, distance, and the curvature of spacetime with extraordinary precision. It is also a reminder that at the frontiers of measurement, the stakes are always high, and there is no margin for getting the units wrong.
The Lesson in All of This
What connects a lost Mars probe, a powerless airliner over Manitoba, a sinking warship in Stockholm harbor, and a Frenchman hiding errors in his notebook is something deceptively simple: measurement is not just about numbers. It is about communication. Every measurement is a message from one person (or one system, or one era) to another, and if the sender and receiver are not speaking the same language of units, the message arrives garbled.
The solutions to these problems are almost always boring. Double-check your units. Label your data clearly. Use standardized systems. Build in redundant checks. None of this is glamorous, and none of it makes for exciting reading in an engineering manual. But the alternative, as these stories show, can be genuinely spectacular in all the wrong ways.
The good news is that awareness of unit-related errors has never been higher. The Mars Climate Orbiter, the Gimli Glider, and the other incidents described here have become teaching staples in engineering, aviation, and medical training programs around the world. Every one of these failures led to better procedures, better checks, and a deeper appreciation for the deceptively simple act of making sure everyone is counting in the same system. The stories survive because they are dramatic and memorable, and because they carry a lesson that never goes out of date: measure twice, convert carefully, and always check which units you are working in.