麻豆传媒

NASA lost its $125 million Mars Climate Orbiter spacecraft as a result
of a mistake that would shame a first-year physics student鈥攆ailing to
convert Imperial units to metric. The problem arose from a culture clash between
spacecraft engineers and navigation specialists, says Mary Hardin, a spokeswoman
for NASA鈥檚 Jet Propulsion Laboratory in Pasadena. 鈥淧ropulsion people talk in
pound-seconds of thrust and navigators talk in newton-seconds,鈥� she says.

The spurious data came from the craft鈥檚 attitude-control system, a design
which had worked fine on the Mars Global Surveyor. But there was one crucial
difference in the system on the orbiter. 鈥淭here was a different propulsion
supplier for the Mars Climate Orbiter, and its data package was in English
[Imperial] units,鈥� says Noel Hinners, vice president for flight systems at
Lockheed Martin Astronautics in Denver, Colorado. No one adapted this
data-processing software for the second probe, so JPL鈥檚 navigation software
thought the numbers it received were newton-seconds rather than pound-seconds.
The attitude thrusters only made small corrections, but the error was enough to
leave the probe 100 kilometres too close to Mars when it tried to enter
orbit.

NASA normally monitors spacecraft positions extensively, but observers
speculate that stressed controllers missed the error. 鈥淓veryone on NASA projects
is incredibly overworked, and mistakes are happening not just because we鈥檙e
faster, but because we鈥檙e working nights and weekends,鈥� says Jonathan McDowell,
an astronomer at the Harvard-Smithsonian Center for Astrophysics in Cambridge,
Massachusetts.

Lockheed, JPL and NASA are now checking for possible data-handling problems
in the ongoing Mars Polar Lander and Stardust comet chaser missions. Hinners
says those probes may be safe because they use different attitude control
systems.