It's been more than three and a half years since the Kepler Space Telescope began its mission as humanity's watcher for Earth-like planets outside of the Solar System. In that time, Kepler has done exactly what was asked of it: provide the data to help identify more than 2,300 exoplanet candidates in other star systems. And so NASA has announced the "successful completion" of Kepler's prime mission. There's one nagging detail, though: we are yet to find a truly Earth-like planet. It's time to alter the parameters of the search.

In the 3 years and 8 months that Kepler has been orbiting the Sun in the Earth's wake, pointing its 95-megapixel camera at over 150,000 stars in the region of the Cygnus, Lyra and Draco constellations, it has identified more than 2,300 exoplanet candidates by recording tiny fluctuations in the brightness of stars as planets pass in front of them over the course of several hours. Though more than a hundred have been confirmed as exoplanets, none yet has been confirmed as truly Earth-like: one of similar size, taking a year to orbit a star also similar to our own.

The same is true for hundreds of other confirmed exoplanets identified by other means, including observatories on the Earth itself. None of the right size are a habitable distance from their stars. What we're looking for is a planet the size of the Earth orbiting a star the size of the Sun with an orbital period of a year: three variables which, if just right, would put the planet in the so-called Goldilocks Zone, at a habitable range.

The criteria may be stringent, but there's one thing in our favor: the sheer number of planets out there. "The initial discoveries of the Kepler mission indicate at least a third of the stars have planets and the number of planets in our galaxy must number in the billions," is the optimistic appraisal of Kepler principal investigator, William Borucki, who suggests that the planet in question may already be hiding among the 2,000 or more candidates already identified by Kepler that are yet to be analyzed.

But Kepler will continue its mission. The reason being that the candidate planets have to be repeatedly analyzed to weed out false positives, which means waiting for them to come back around for another orbit. "Because of the high variability of stars like the sun, we're not going to get the final answers in four years," Borucki told Spaceflight Now in October. "We're going to get them in eight years." If the last three and half years were about identifying possible Earth-like planets, the next four will be about scrutinizing them.

There is also the problem of noise interference, which has required the development of elaborate computer processing to address that must recalibrate every last pixel in an image.

Though an extension of several years, perhaps as far as 2016, was announced in April, the latest confirmation of Kepler's continuation is welcome news. The extended mission's longevity was threatened in July when one of Kepler's four reaction wheels, which are used to fine-tune its orientation, ceased. Though Kepler can continue to function, the remaining three must continue to work in order to do so. "The guiding is still great, but they've all had over a billion revolutions," Borucki told Spaceflight Now. "If we lose another one, this mission terminates. We cannot track very well with two. We cannot track well enough to find planets."

Kepler has cast its net very wide indeed since 2009. Between now and 2016 and beyond, it's a case of gradually pulling it in.

Sources: NASA, Spaceflight Now