Updated 5:42 PM ET
NASA's $600 million Kepler Space Telescope, a leading player in the hunt for Earth-like planets orbiting sun-like stars, has been sidelined by problems with stabilizing gyroscopic reaction wheels needed to hold the spacecraft and its 95-megapixel camera precisely on target, NASA managers said Wednesday.
One of Kepler's four reaction wheels was shut down last July and problems with a second wheel -- No. 4 -- were discovered Tuesday when engineers found the spacecraft had put itself into a protective "safe mode" after running into a problem of some sort. Subsequent commands to spin up the No. 4 wheel were not successful.
While the space telescope is equipped with rocket thrusters for major maneuvers, at least three reaction wheels are needed to precisely orient the spacecraft and keep it accurately aimed at its target starfield.
"We need three wheels in service to give us the pointing precision that's necessary for us to find planets," project scientist William Borucki told reporters. "That's what makes this mission work, the fact that we can point with extreme precision, as well as the Hubble Space Telescope, in fact. So without three wheels, it's unclear whether we can do anything of that order."
Engineers have not yet given up hope and they are studying options for restarting one of the stalled wheels or developing techniques for operating the spacecraft with just two.
In the meantime, Kepler's search for exoplanets is on hold. Commands were uplinked to put the spacecraft in a configuration to conserve rocket fuel and give engineers time to thoroughly study the problem and possibly come up with a fix.
"There is a reasonable possibility we will be able to mitigate that problem," Borucki said during an afternoon teleconference. "So I don't think I'd be a pessimist here. The mission's been phenomenally successful, and I really wouldn't write it off yet."
But any repair will have to be accomplished by remote control. Kepler is trailing Earth in its orbit around the sun and currently is about 40 million miles away, far beyond the reach of astronauts in low-Earth orbit.
Launched March 6, 2009, the Kepler space telescope is equipped with a 95-megapixel camera designed to continually monitor the light from more than 100,000 stars in a patch of sky in the constellation Lyra.
The camera acts as a photometer of sorts, continually monitoring the brightness of candidate stars in its wide field of view and the slight dimming expected to result if planets happen to pass between target stars and the telescope.
It was a technological challenge, comparable to watching a flea creep across a car's headlight at night. But by timing repeated cycles, computer analysis can tease out potential Earth-like worlds in habitable-zone orbits.
The probability of finding sun-like stars with Earth-like planets in orbits similar to ours -- and aligned so that Kepler could "see" them -- is only about one-half of 1 percent. But given the number of stars in question, however, that still leaves a sizeable population of potential candidates.
Going into the mission, scientists said it would take about four years of around-the-clock observations to capture the repeated cycles needed to confirm detection of an Earth-like world in an Earth-like orbit, one in the parent star's habitable zone where water can exist as a liquid.
NASA's Exoplanet Archive lists 866 confirmed planets orbiting 671 stars, including 133 solar systems hosting multiple planets.
Kepler currently is credited with discovery of 132 of those confirmed planets and 2,165 planet candidates that require additional analysis and observation. Despite the reaction wheel problem, scientists expect a steady stream of discoveries in the months ahead as additional analysis of archived data allows them to move candidate planets into the "confirmed" column.
"The mission was designed for four years, it operated four years, it gave us excellent data for four years," Borucki said. "So I'm very delighted.
"On the other hand, I would have been even happier if it had continued another four years because we'd have better data about more stars, about smaller planets and we'd have more planets that we'd probably find in the habitable zone. So that would have been frosting on the cake. But we have an excellent cake right now."