NASA's latest robotic mission to detect life will get under way this week as Carnegie Mellon University researchers pack up the robot known as Zoe and ship it to Chile's Atacama Desert.

	Pam Panchak, Post-Gazette Dan Villa, a second-year master's-degree student at Carnegie Mellon University's Robotics Institute, takes Zoe out on Flagstaff Hill in Oakland earlier this summer in a test of the robot's positioning system. Click photo for larger image. For more information, visit www.frc.ri.cmu.edu/atacama/

Zoe will spend a week exploring each of three "landing sites" within this driest of deserts, searching for signs of life that can be so minute that they are undetectable to the human eye.

The two-month-long field trial, this project's third visit to the Atacama, will test instruments and techniques that might someday be used to search for life on Mars. Finding evidence of microscopic life in extreme environments can be tricky and no one yet knows quite how to do it with remote sensing.

"This is an extremely important experiment this year," said Nathalie Cabrol, a planetary scientist at NASA Ames Research Center and the leader of the science team for the Life in the Atacama project. Though Zoe became the first robot to remotely detect life during last year's field work in the Atacama, Cabrol hopes that the evidence of life gathered this year will be unambiguous.

"Last year was very encouraging," she said, noting that the "hints of life" in the visual evidence gathered by Zoe were ultimately confirmed by soil and rock samples that were retrieved and are still being studied in the laboratory. "But we ask ourselves, if Zoe had been on Mars and we didn't have sample return, could we be positive?"

In some cases, Zoe detected lichens, which were reasonably easy to identify. But in other cases, particularly in the driest areas of the desert, the only life the robot uncovered was microscopic bacteria.

Zoe, a solar-powered robot that sits on bicycle wheels, was built at CMU's Field Robotics Center and its main instrument, a fluorescent imaging camera, was developed by the university's Molecular Biosensor and Imaging Center.

The instrument looks for the natural fluorescence of organisms that contain chlorophyll and also uses fluorescent dyes that bind to DNA, protein, carbohydrates or lipids.

Last year, researchers had to reach under the robot and spray rocks and soil with the dyes, but the Robotics Institute outfitted the instrument with an automatic sprayer for this field test.

"A human shouldn't have to get anywhere close to the robot," said Alan Waggoner, director of the biosensor center and one of the researchers who did the manual spraying last year. But that will pose new challenges, such as how to figure out if the sprayer becomes clogged.

David Wettergreen, project leader and an associate research professor at the Robotics Institute, said the robot itself should require little attention this year and the robot and its instruments are now fully integrated, so no human intervention will be necessary.

A field team will continue to gather samples of soil and rocks imaged by Zoe, but it won't be following the robot around. Instead, Zoe will mark each spot it images with a couple of dots of a marking solution; guided by GPS coordinates, the tracks made by Zoe's wheels and the painted dots, the researchers can gather the samples a day later, said biologist Shmuel Weinstein.

Following each day of exploration, Zoe will transmit images to a multidisciplinary science team in Pittsburgh, which will evaluate the images to see if they are likely to contain life. The team, which will include biologists, geologists and others, will send instructions to Zoe each night, suggesting areas that should be explored the next day.

From a tactical standpoint, Cabrol said she hopes that the team will be unafraid to backtrack at times. During last year's field trial, she noted, science team members often found themselves saying, "Oh, it would be great if [the robot] could go back there." So she hopes to occasionally take advantage of hindsight and send Zoe back for some second looks.

Waggoner said the researchers also have learned from previous field trials that simply spraying a little water on the spot being studied is enough to quickly rejuvenate any dormant life. A spritz of water can sometimes be enough to cause natural fluorescence.

And any change in fluorescence detected after a spritz of water can itself be a sign of life, he added. "If it was mineral fluorescence," he explained, "it probably wouldn't change when you sprayed water on them."