A new world in your bedroom

IN AN age of professionals, the ability of amateur scientists to make meaningful contributions has almost vanished. Almost, but not quite. The internet allows professionals to make their data available for analysis by anyone, and some are happy to take advantage of the free labour this promises. This approach has proved particularly fruitful in astronomy, a science with a long history of amateur contributions. Armchair astronomers have already helped classify galaxies seen by Hubble, the main orbiting telescope of America's space agency, NASA. They have also looked for interesting asteroids, and kept an eye out for solar storms.

The latest project to involve them, called Planet Hunters, allows amateurs to search for extrasolar planets—those that orbit stars other than the sun. It was set up by a group at the universities of Oxford and Yale, and links 40,000 participants with data gathered by Kepler, another NASA space telescope that is specifically designed to hunt for planets. On September 26th the group announced, in a paper posted to arXiv, an online database, that its participants had discovered two probable exoplanets, one a Jupiter-like gas giant, and the other, possibly, a smaller, rocky world about twice the diameter of Earth.

Kepler works by monitoring the thousands of stars in its field of view for tiny changes in brightness. Mostly, these are natural fluctuations, but particularly sharp and regular changes might signify a planet passing in front of a star. The raw data are sent to computers on Earth, converted into graphical form and made available to the Planet Hunters. After logging onto the project's website, its users are given a brightness graph from a random star and asked to mark anything of interest. If several people flag the same star, the result is checked against the computer-derived results produced by the main Kepler team. Promising candidates are then checked again by ground-based telescopes.

That allows Planet Hunters' participants both to act as a benchmark for the star-detection algorithms and to discover planets the computers have missed, says Chris Lintott, an astronomer at Oxford who helps to run the project. What people lack in speed (the computers have already notched up over 1,200 candidate planets since Kepler was launched) they make up for in judgment. Some stars being watched have very variable brightness. That is confusing for computers, but for human eyes is less of a problem. And input from the human planet hunters is used to refine the algorithms, improving their performance.

Planet Hunters grew out of Galaxy Zoo, which was set up in 2007 to help researchers classify galaxies spotted by Hubble—just the sort of fuzzy task that machines struggle with but humans excel at. Galaxy Zoo spawned the Zooniverse, a collection of science projects that harness the power of amateurs. Although astronomical projects still dominate, other sciences are starting to adopt the idea. One Zooniverse project aims to reconstruct weather records from old Navy logs; another is helping to transcribe a cache of Egyptian papyri dating from the 1st century AD. Dr Lintott and his colleagues have asked researchers in other fields to submit more ideas, and hope to announce the shortlist in a few weeks' time.