mars climate orbiter became a cautionary NASA case after the accidental use of Imperial instead of metric units, a breakdown that ultimately meant doom for the spacecraft. The mission launched on December 11, 1998, alongside the Polar Lander, under NASA’s broader Mars Surveyor ’98 initiative, with Lockheed Martin commissioned to design and build the orbiter and NASA’s Jet Propulsion Laboratory overseeing the project. By the morning of September 23, 1999 (ET), communication was gone and could not be reestablished, ending the effort before the spacecraft could begin its planned work gathering Martian weather data while supporting communications with the Polar Lander.
What failed, and when NASA realized something was wrong
On paper, the project had layered oversight and a clear flight plan. The spacecraft was equipped with eight thrusters intended to boost it into Mars orbit, plus reaction wheels used to adjust altitude and orientation. Those reaction wheels sometimes “overdid the momentum, ” triggering the need for an angular momentum desaturation (AMD) event to reset the system—procedures that relied on reliable navigation calculations on Earth.
But soon after launch, software problems began to surface. During the trip—projected to last nine and a half months—the spacecraft’s software began acting up, and ground navigation data had to be emailed to NASA for solutions. Even after corrections, the spacecraft was still transmitting “nonsensical data” back to Earth, raising the stakes for each navigation decision as the approach to Mars tightened.
In September 1999, engineers computed and executed the final planned Trajectory Correction Maneuver, known as TCM-4, to refine the approach. The intended outcome would have produced a first periapsis of about 140 miles (226 km) above Mars after orbit insertion. Navigators then determined the predicted closest approach was lower than expected, exposing a serious trajectory error as Mars’ gravity began pulling the orbiter in.
NASA postmortem: the Imperial-vs-metric mismatch at the center
By the morning of September 23, 1999 (ET), the spacecraft had vanished, leaving teams with no way to reestablish communication. NASA’s initial postmortem analysis concluded the spacecraft was only about 35 miles (57 km) above the Martian surface when contact was lost. Engineers concluded it either burned up in the Martian atmosphere or skipped off the atmosphere and was lost in space.
When the agency investigated the following month and identified a data issue, attention focused on the “small forces” software used to determine position and plan AMD events. Investigators found a critical inconsistency: while everything else used metric units, this software was using Imperial units. Lockheed Martin’s use of the wrong units in software meant the spacecraft was not close to the trajectory it was supposed to be on.
NASA had required Lockheed Martin to convert measurements to metric units, but the agency did not verify which measurement system the company had employed before sending the spacecraft toward Mars. The investigation also noted that navigation staff raised concerns during the mission and there was reportedly no response from upper management.
Immediate reactions and responsibility findings
NASA’s initial postmortem analysis laid out the mission’s last known conditions near Mars and tied the failure to the units mismatch embedded in flight-related software. Investigators also claimed NASA was responsible for the mission failure rather than Lockheed Martin, pointing to breakdowns in verification and management response as the spacecraft continued transmitting faulty information and navigators flagged concerns.
At the heart of the case was a basic systems engineering failure: metric was required, Imperial appeared in a key software component, and the mismatch was not caught before the final approach.
Quick context and what comes next
The spacecraft was built to gather data on Martian weather and communicate with the Polar Lander as part of Mars Surveyor ’98, but it never reached operational orbit. The mission’s end came as trajectory predictions tightened and the computed closest approach dropped below expectations.
Next steps in cases like this typically hinge on what the postmortem establishes: how the verification failed, how software interfaces were checked, and why warnings did not trigger intervention. For the mars climate orbiter, the record in NASA’s initial postmortem analysis makes the central lesson unmistakable: a units mismatch inside mission-critical software can erase even extensive oversight—right when a spacecraft is seconds and miles from the target world.
