Mars Climate Orbiter: Difference between revisions
| (page creation) | Newer revision (05:25, 28 September 2026) 52.38.225.52 (New page from the article seed) |
| Line 1: | Line 1: | ||
| + | {{Infobox | ||
| + | | title = Mars Climate Orbiter | ||
| + | | operator = NASA, with Lockheed Martin | ||
| + | | launched = 11 December 1998 | ||
| + | | lost = 23 September 1999 | ||
| + | | root cause = a ground software file reported pound-seconds where the specification required newton-seconds | ||
| + | | factor = 4.45 | ||
| + | | software on the spacecraft = metric, and correct | ||
| + | }} | ||
| + | The '''Mars Climate Orbiter''' was a NASA spacecraft lost on 23 September 1999 as it arrived at Mars, after a nine-month flight. The onboard software used metric units and was working. A file produced on the ground, by a program called SM_FORCES, reported the impulse of small thruster firings in pound-force seconds. The specification said newton-seconds. Navigation software believed the specification.<ref name="mib">Arthur G. Stephenson and others, ''Mars Climate Orbiter Mishap Investigation Board Phase I Report'', NASA, 10 November 1999.</ref> | ||
| + | |||
| + | One pound-force is 4.45 newtons. The navigation solution was therefore low by 4.45, for the whole cruise. | ||
| + | |||
| + | == What the board found == | ||
| + | The Mishap Investigation Board's root cause is a single sentence. The ground software did not use metric units in the "Small Forces" file used to model the trajectory. The Angular Momentum Desaturation file was supposed to be in newton-seconds. It was in pound-seconds. Every later calculation of where the spacecraft was, built on that file, undercounted the effect of the thrusters by the conversion between the two.<ref name="mib" /> | ||
| + | |||
| + | The spacecraft itself had been computing the same forces in metric units, correctly. The ground did not use those numbers. The error was found on 29 September, six days after the spacecraft had already been sent into the Martian atmosphere on a path it could not survive.<ref name="mib" /> | ||
| + | |||
| + | The board did not stop at the units. It listed the things that let a units error travel for nine months: the mission was not treated as one system, the interface specification was not followed and was not checked end to end, and a last navigation fix before arrival was considered and not done. Arthur Stephenson, the board's chairman, said the failed translation of units was the root cause, and that other failures had let it "linger and propagate".<ref name="nasa">"Mars Climate Orbiter Failure Board Releases Report", NASA, 10 November 1999.</ref> | ||
| + | |||
| + | The number 4.45 is not a large number. Multiplied by every thruster firing between Earth and Mars, it was the difference between an orbit and a hole in the atmosphere. | ||
| + | |||
| + | == See also == | ||
| + | * [[Gimli Glider]] | ||
| + | |||
| + | == References == | ||
| + | {{reflist}} | ||
| + | |||
| + | [[Category:True but improbable]] | ||
| + | |||