Gentlemen, start your gyros!
NASA's Gravity Probe B spacecraft has begun its search for a bizarre prediction of Einstein's relativity.
September 2, 2004: It's "all systems go" for one of the most ambitious physics experiments ever attempted.
Right: Gravity Probe B in Earth orbit. 
"It's a long and tortuous story," says Francis Everitt, principal investigator for Gravity Probe B (GP-B) and a professor at Stanford University.
Instead it took weeks.
First, sunlight reflecting off floating dust particles confused the satellite's star-tracking sensors. These sensors use the locations of constellations to orient the spacecraft, and the tiny shining specs looked like stars. The dust eventually cleared, but then another problem arose: Cosmic radiation in the form of high-speed protons peppered the telescope's light sensor, causing false signals. Mission scientists had to tweak the satellite's software to ignore these pulses. And on it went like this for weeks; scientists would solve one problem only to encounter another.
"Now it has become very routine, and we only take about a minute to acquire the star as we come up over the horizon," Everitt says. (The satellite loses sight of the guide star during each orbit because it passes behind the Earth, so it must reacquire the star as it comes back into sight.)
Above: A spinning spherical gyroscope in Earth orbit should wobble in the whirlpool shaped space-time around our planet. [More]
"One of the things all of us were terribly worried about was getting some dirt in the gyro housings," Everitt says. The gyros float a near-perfect vacuum, and only a thousandth-of-an-inch gap separates the spheres from their casings.
"The gyros were cleaned before they went up, but we gave this thing a tremendous vibration during launch. Wouldn't you expect a piece of dirt to come in through one of the pump-out ports, land right on one of the gyros and jam it?" he says. "That would be the end of that gyro."
This time all the worrying was for nothing. "The gyros have all been as clean as a whistle," he says. They're suspended in their casings, aligned with the guide star, and spinning thousands of times per minute. "Amazing, delightful."
Above: Gravity Probe B mission control at Stanford University.
Now the gathering of science data begins. The satellite's onboard computers should be able to handle this phase of the mission automatically. Still, at least one person will be on duty monitoring GP-B at all times throughout the year, Everitt says. "It should run itself, but you can never relax."
After more than 40 years of methodical planning and four months of intense troubleshooting, GP-B's scientists feel "a real sense of gladness," he says. "What a difference it makes to be up there and operating. How thrilling that is. We all feel that."
"Some people," laughs Everitt, "are talking about taking a week or two of well-deserved vacation."