The Time NASA Lost a Mars Orbiter Because of a Metric System Mixup

Using the metric system onboard a spacecraft and imperial on the ground can have disastrous consequences.

From the Viking landers to our now more than ten years of continual roving on the red planet, NASA has had a lot of success exploring Mars. But one instance sticks out as an unfortunate blemish on NASA's Martian record: the loss of the Mars Climate Orbiter, which was chalked up to an unfortunate mixup between imperial and metric units of measurement. 

So how exactly did engineers get their units confused? To highlight the 15th anniversary of MCO's disappearance, let's look back at just how NASA lost a spacecraft.

The story of the Mars Climate Orbiter starts in 1993, when NASA initiated the Mars Surveyor program. The goal of this multi-mission program, which was based at NASA's Jet Propulsion Laboratory, was to explore the Martian climate in far more detail than the Viking probes were capable of. 

In 1995, two missions were conceived to take advantage of a favorable launch window that spanned the end of 1998 to the beginning of 1999: the Mars Climate Orbiter and the Mars Polar Lander. As their names suggest, MCO was designed to orbit Mars, serving as both the first interplanetary weather satellite as well as a communications relay for MPL.

An artist's rendering of the MCO. Image: JPL

MCO launched on December 11, 1998 atop a Delta II rocket from Launch Complex 17A at the Cape Canaveral Air Station in Florida. The spacecraft left the Earth on a trajectory such that it would take nine months to arrive at Mars, but they wouldn't be boring months.

You can't just launch a spacecraft towards Mars and let it go, trusting it will arrive at the right place at the right time. Mission scientists have to manage the spacecraft constantly to make sure its stays on the right path. Spacecraft like MCO have reaction wheels—spinning units that keep it oriented in a certain way—and the spacecraft's angular momentum must be constantly monitored and adjusted to keep those reaction wheels working within range. And this is where all the trouble started.

Remember, we're talking about distances of millions of kilometers, which means adjusting a spacecraft's angular momentum and reaction wheels isn't the most simple process.

First, spacecraft data is transferred to the ground and processed by a software program; in the case of MCO the program was called SM_FORCES (small forces). The data is processed then placed into an Angular Momentum Desaturation file. The operations navigation team from JPL then used the information in that file to model the forces acting on the spacecraft and adjusted its trajectory by firing its thrusters as needed. 

After every thruster firing, the change in velocity was measured and calculations about its performance carried out both by the spacecraft and by support computers on Earth.

The MCO team didn't start using those ground computers' data to adjust the spacecraft's trajectory until four months into the mission, but almost immediately it was apparent that something wasn't working right. The numbers from the spacecraft and the ground computers didn't match.

It turned out that the two systems were using different units of measurements. The Software Interface Specification that was used to define the Angular Momentum Desaturation file on board the spacecraft relied on the widely used—and metric—Newton-seconds as the units for measuring impulse (which is defined as force multiplied by time). 

The ground computers, on the other hand, used software that used imperial pound-seconds to measure impulse. So metric onboard the spacecraft, imperial on the ground. 

This means that every impulse measurement done by the Angular Momentum Desaturation software underestimated the effects of the thrusters on the spacecraft by a factor of 4.45 (1 pound of force is equal to 4.45 Newtons).

When you're traveling tens of millions of kilometers, every little deviation adds up. The measurement errors in the ground-based processing went overlooked by the ground crews; modelers assumed that the data provided to the spacecraft was in the correct metric units.

Compounding the problem was the fact that MCO was an asymmetrical spacecraft. Its solar array sat asymmetrically relative to its body, so the effects of the pressure from the solar wind was intensified. This meant the spacecraft had to make more corrective maneuvers to counteract the effects of the solar wind, 10-14 times the number mission planners had originally anticipated performing.

Nevertheless, the spacecraft made it to Mars in September of 1999, nine and a half months after launch. Once there, MCO was set to fire its main engine for the orbit insertion burn that would put it into an elliptical orbit around Mars. It was then supposed to go through a period of aerobraking, skimming though the upper atmosphere for weeks to gradually lose speed. 

But it didn't. After that orbit insertion burn, NASA abruptly lost contact with the spacecraft.

Throughout the nine month cruise to Mars, seven small errors were introduced into MCO's trajectory estimates. Together, these errors put the spacecraft 105 miles (170 kilometers) closer to the Martian surface than it should have been when it encountered the planet. It was, quite simply, an unsurvivably low encounter altitude.

And unfortunately, we may never know what exactly happened. At that low encounter altitude, MCO was either destroyed as it entered the Martian atmosphere or passed through the Martian atmosphere and settled into orbit around the Sun.

Regardless of the spacecraft's fate, losing MCO was an embarrassing blow to Martian exploration. The silver lining, albeit a thin one, is that we learned a valuable lesson: double checking units between systems that have to work together is never worth overlooking. Because, really, you'd think rocket scientists would know better. But hey, live and learn.