NASA Auditions Robots for Lunar Exploration Missions

Arizona's famous Meteor Crater is a long way from the Moon.But for a menagerie of intelligent robots hoping to earn supporting roles inNASA's lunar exploration plans, the massive impact crater east of Flagstaff iscenter stage.

In September, several such robots and an autonomous Moonbuggy called Scout were put through their paces in the rough desert terrain.During a two-week campaign conducted by NASA's Desert Research and TechnologyStudies team -- a collection of government, university and industry scientistsand engineers known as the Desert Rats -- the robots demonstrated their abilityto work side-by-side with space-suited researchers, helping with the kinds oftasks that actual astronauts will have to perform as they begin exploring theMoon and establishing outposts.

NASA's current expenditure on so-called robotic fieldassistants is fairly modest. Of the $3 billion NASA's Exploration SystemsMission Directorate spent this year developing a new space transportationsystem and preparing for an eventual return to the Moon, only $13 million wentto Human Robotics Systems, a recently established program meant to focus theagency's investment in robotic helpmates.

Chris Culbert, the NASA Johnson Space Center engineer incharge of the Human Robotic Systems program, said his current stable ofprototype field assistants consists almost entirely of robots inherited fromvarious programs around the agency, some of which pre-dated President George W.Bush's 2004 call for the United States to return to the Moon.

"There's not enough money in this program and it hasn't beenaround long enough to build up new robots," Culbert said.

In addition to Scout, NASA's current line up of fieldassistants includes a nimble six-legged rover called Athlete, a dexterous humanoidtorso on wheels called Centaur, and K-10, a boxy little rover speciallyequipped for site survey work.

Culbert said all four robots help NASA in one way or anotherto address the three big themes of the Human Robotics Systems program: surfacemobility, surface handling, and human-systems interaction.

"The interaction between robots and humans is very importantto me," Culbert said. "Industrial robots are typically behind barriers and bigalarms ring if humans come within 10 feet. Our robots live with the humans."

Athlete and Centaur, two of the robots that were used atMeteor Crater this fall, will be on display and demonstrated at NASA's 2ndExploration Conference being held Dec. 4-6 in Houston.

When Centaur rolls into the exhibit hall at the George C. Brown Convention Center, it will be a sort of homecoming for the wheeled robot'shumanoid half, a skilled construction worker formerly known as Robonaut. NASAoriginally developed Robonaut with an eye toward helping astronauts withtedious tasks like those they were encountering assembling the internationalspace station. At the 2002 World Space Congress in Houston, a stationary Robonautwas on exhibit showing off its dexterity with hand tools. NASA officialspresent there talked about how Robonaut could one day help assemble very largeaperture space telescopes in orbit.

Since mounting Robonaut on a wheeled platform and renaminghim Centaur after the half man/half-horse of Greek mythology, NASA has devotedmore attention to thinking about how the dexterous robot might help build andmaintain outposts on the Moon. Culbert said a robot like Centaur might make anexcellent lunar plumber, for example, tackling various pipefitting chores andfreeing astronauts to concentrate on more important activities. But Centaur canalso do grunt work. In the desert in September, Culbert said, the robot helpedthe stand-in astronauts unload the Scout rover.

Athlete got its start in late 2004 when NASA was spendingrelatively freely on human and robotic technology projects that supported awide mix near and longer-term exploration goals.

NASA initially planned to spend $25 million on Athlete overfour years. But after NASA Administrator Mike Griffin took over in early 2005and decided the agency could not afford the $1 billion-a-year technologyportfolio his predecessor had set in motion, all but 38 of the original 118competitively-selected projects were canceled. Athlete made the cut, but hadits funding reduced substantially. After receiving $3 million in its first year,Athlete's funding was paired back to $1.5 million.

Brian Wilcox, the Jet Propulsion Laboratory-based principleinvestigator for Athlete, said the funding reduction ruled out building ahigh-fidelity prototype out of mission-grade components as the team originallyproposed. But the team has managed to build three somewhat lower fidelityAthlete prototypes, including two fully-functional vehicles, fromcommercially-available components. The semi-autonomous rovers run softwarebrought over from the Mars Exploration Rover program.

Wilcox said Athlete is first and foremost a versatile cargovehicle that has demonstrated its ability to tackle a wide range of tasks andterrain challenges, including climbing hills and rappelling into ravines.

Athlete also would make a great set of landing legs forNASA's next human lunar lander, Wilcox said. "We have been promoting theconcept that Athlete would make superb landing legs for the [Lunar SurfaceAccess Module]," he said. "It would give the first lander superb mobility aswell as a general purpose manipulator, so it would have the ability to do selfmaintenance."

What the future holds for Athlete and the rest of itsrobotic brethren depends in part on how NASA's lunar exploration plans takeshape in the years ahead. NASA is due to release its first stab at a lunarsurface architecture during the Exploration Conference in Houston.

Culbert said NASA has more work to do before it can saywhether six-legged rovers like Athlete are the answer to its lunar surfacemobility needs, or if a "simple flatbed truck" approach might be a better wayto go. "We may find out outpost on the Moon doesn't require a lot of climbingup and down hills," he said.

To help answer these questions, Culbert and his team plan tokeep building robots and putting them through their paces. But he admits thatthese important early efforts are for the time being rather resourceconstrained.

"With the amount of money I've got available I won't be ableto do everything that needs to be done," he said.

NASA hopes to spend more money on Human Robotic Systems inthe years ahead. NASA's 2007 budget request, still awaiting approval fromCongress, includes $18 million for Culbert and the team of 40 or so civilservants and contractor personnel spread across the agency. NASA's most recentfive-year-plan forecast ramping the program up to $25 billion a year by 2011,but Culbert admitted that might be a tough sell given some of the otherpressures on the Exploration Systems Mission Directorate's budget.

If additional money is forthcoming, Culbert said the programwould like to build a new moon buggy next year equipped with an activesuspension system tuned to handle tougher terrain. Human Robotic Systems groupmay also build a crane in 2007 and possibly add it to one of the Athleterovers.