Bound for the moon

The next rover to roam the moon’s surface

may come not from nasa and its rocket scientists

but from college students and

private companies working on a shoestring

By Michael Belfiore

On a muddy, rubble-strewn field on the banks of the Monongahela River in Pittsburgh, a five-foot-tall pyramidal robot with twin camera eyes slowly rotates on four metal wheels, its electric motors emitting a low whine. In a nearby trailer, students from Carnegie Mellon University huddle around a laptop to watch the world through the robot’s eyes. In the low-resolution grayscale images on the laptop’s screen, the rutted landscape looks a lot like the moon, which is the robot’s ultimate destination. Carnegie Mellon robotics professor William “Red” Whittaker and his students built Red Rover to win the Google Lunar X PRIZE, a competition designed to boost the role of private companies in space and inspire innovation in spaceflight technology. The winning prize is $20 million, which will go to the first nongovernment team that lands a robot on the moon, gets the robot to travel half a mile or so, and sends high-definition video back to Earth—all by the end of 2015. A second-place prize of $5 million, along with bonuses for other achievements such as reaching the site of an Apollo landing, brings the total purse to $30 million. Although 26 teams are competing, Whittaker’s team is a clear leader. His firm, Astrobotic Technology, was the first team to make a down payment on a rocket that will carry its spacecraft and rover to the moon. Whittaker has also proved himself to be a champion builder of autonomous vehicles that can navigate extreme environments. The Google Lunar X PRIZE comes at a major turning point for the U.S. space program. In 2010, following the recommendations of the Review of U.S. Human Space Flight Plans Committee, President Barack Obama directed NASA to encourage privately owned and operated spaceships to replace the retiring space shuttle. With input and seed money from NASA, the reasoning goes, private companies can design and construct ships more quickly and more affordably than the usual big contractors can produce vehicles for the government agency. In the same spirit, the Google Lunar X PRIZE seeks to foster a new class of private planetary missions, one that does not depend on expensive one-off spacecraft and political commitments that may not last beyond one administration. Instead researchers would pay private companies to launch their rovers and instruments. NASA has added its own incentives—an additional $30.1 million, split among six teams for surmounting technical feats that have stumped many government rovers, such as surviving the lunar night. The fate of private spaceflight companies after the Google Lunar X PRIZE is far from certain, and not everyone is convinced that a market exists for their services, but many researchers are excited about the prospect of commercially funded space science. TEST LAUNCH the contest has a precedent in the $10-million Ansari X PRIZE, which ended in 2004, when SpaceShipOne became the first privately manufactured manned vehicle to leave the atmosphere. SpaceShipOne was a rocket plane built by Mojave, Calif.–based Scaled Composites, with funding from Microsoft billionaire Paul Allen. Virgin Galactic is now financing SpaceShipTwo. It has received more than $60 million in deposits from individuals who are willing to pay $200,000 each for the chance to float in microgravity and see Earth from a distance. NASA has contracted Virgin and six other private companies to fly scientific equipment onboard SpaceShipTwo and other spacecraft to conduct experiments on challenges such as transferring fuel without gravity. Now the organizers of the Google Lunar X PRIZE hope to duplicate this success for robotic planetary missions. Few people are as qualified to get a robot on the moon as Red Whittaker. The 63-year-old may have done more than any other individual in developing the discipline of field robotics— taking robots out of controlled environments such as automobile factories and releasing them to do useful work in the wild. In the 1980s he designed and built the robots that explored damaged and dangerously radioactive areas of the partially melted-down Three Mile Island nuclear power plant. As founder and head of the Field Robotics Center at Carnegie Mellon, Whittaker has since made a career of breaking new ground in autonomous vehicles. He has created robots that hunt meteorites in the ice fields of Antarctica and robots that climb into the craters of active volcanoes in Alaska and Antarctica. Whittaker began planning for the Google Lunar X PRIZE in 2007 while in the midst of a different competition: the Defense Advanced Research Projects Agency’s Urban Challenge, held at the former George Air Force Base in Victorville, Calif. Under the team name “Tartan Racing,” Whittaker and his students partnered with General Motors, Continental and other sponsors to create a driverless Chevy Tahoe named “Boss.” Even as he won a first-place victory in the world’s first autonomous vehicle race through city streets, Whittaker wasted no time in finalizing plans for a class at Carnegie Mellon called Advanced Mobile Robot Development. The class’s modest objectives, as described in the course catalogue, are to “detail, analyze and simulate a robotic lunar lander, field-test a lunar rover prototype, tackle enterprise challenges, and communicate mission progress through writing, photography and video.” The course is open to Carnegie Mellon students of any field at any level. Around the same time, Whittaker established Astrobotic Technology as a for-profit company with long-time space entrepreneur David Gump at the helm. Gump aggressively pursues corporate sponsorships and potential customers, whereas Whittaker contributes deep knowledge accumulated over more than 29 years of research at the Field Robotics Center. Among Astrobotic’s sponsors is Pittsburgh-based Alcoa, which has donated the aluminum required for the spacecraft that will carry the rover to the moon. Whittaker, an ex-marine and the son of a chemist and an explosives salesman, says that landing one of his team’s creations on the moon would represent the fulfillment of a career path that has seen his robots on land, water, underwater, underground, and in just about every environmental extreme here on Earth. Winning the moon doesn’t just mean the first prize; in his mind, Astrobotic won’t be successful until it meets every one of the bonus objectives as well. “If you haven’t done everything,” he says, “you haven’t done anything.” ROCKET SCIENCE whittaker’s vision for getting Astrobotic’s spacecraft and rover on the moon begins with the SpaceX Falcon 9 rocket. Established with the goal of dramatically reducing the cost of space access, SpaceX may be the key enabler of the Google Lunar X PRIZE competition. Whittaker believes that the SpaceX rocket will be the vehicle of choice for all the teams in the competition. “As far as I’m aware, every U.S. contender is targeting SpaceX,” he says. Even so, the cost of launch will be the single greatest expense for any team. Though less expensive than other rockets in its class, the published price of a Falcon 9 launch is still $54 million— more than twice the top prize. SpaceX’s competitors are reluctant to discuss their own launch arrangements, but it is clear that SpaceX has already upended the market with the single biggest commercial launch contract in history—a $492-million deal with Iridium, a satellite communications company. After Red Rover leaves Earth’s atmosphere atop its Falcon 9, the Astrobotic spacecraft-and-rover stack will jettison its protective nose fairing, and the rocket’s second-stage engine will push the spacecraft and rover on a course to the moon. The transit will take five days. Guidance, navigation and control software developed at Carnegie Mellon will keep the rocket on the right path. The software is a direct descendant of the code that enabled Tartan Racing to win the Urban Challenge. The computational challenges of autonomous driving and spacecraft piloting are not so different—the same kind of math solves both problems, which is why the software is so similar. The main difference, says Astrobotic team member and Ph.D. candidate Kevin Peterson, is the lack of GPS to guide the vehicle. Instead the craft will plot its trajectory to the moon by referencing stars, the moon and Earth. Once in orbit, the spacecraft and rover must descend to the moon’s surface. In 1969 astronaut Neil Armstrong piloted the lunar module from orbit to a specific location on the moon, while avoiding local hazards such as boulders and craters. But the 250,000-mile distance between our planet and its satellite imposes a time lag that precludes real-time control by a pilot on Earth, so the spacecraft’s software will have to accomplish autonomously what Armstrong did by hand. A primary descent engine will burn to slow the spacecraft down as it approaches the moon, while small thrusters will keep the vehicle stabilized. Touching down two days after lunar dawn, the lander will deploy two ramps (the second is a spare, in case a rock or crater obstructs the first). The bolts that hold the ramps folded against the ladder are rigged to break apart under intense heat. After the ramps fall from the spacecraft to the ground, the rover will roll down one of them to the moon’s surface, binocular eyes scanning the ground ahead. Moon dust is too slippery to permit an accurate reading of distance traveled based on how many times the rover’s wheels have turned. Instead the rover’s onboard computer will calculate distance by comparing the changing appearance of surface features as the robot moves. Radiation- hardened components will protect the computer from the unfiltered solar and cosmic radiation with which the airless moon is bombarded. Back in Pittsburgh, Astrobotic team members at mission control will work 24-hour shifts through the long lunar day, using a steady stream of low-resolution images to guide Red Rover to interesting features (including, it is hoped, an Apollo landing site). The rover will avoid hazards on the moon’s surface autonomously. It will beam high-definition video as blocks of encrypted data, at least one immediately after landing and one later in the mission to meet X PRIZE requirements. The rover will also send e-mail, tweets and Facebook posts. A major technical challenge for the team is making sure Red Rover survives the extremes of lunar day and night, each of which lasts two Earth weeks. During the two-week lunar night, the temperature at the moon’s surface where the team plans to land plummets from a daytime high above 248 degrees Fahrenheit to around −274 degrees F. Any components that contained traces of water, such as the batteries, would suffer irreparable damage as the water froze and expanded. The only rovers ever to have survived the extremes of day and night were the Soviet remote-controlled lunar rovers, called Lunokhods, in the 1970s. They relied on a radioactive polonium isotope to stay warm. But Astrobotic and other private companies competing for the X PRIZE do not have access to these tightly controlled materials. To protect Red Rover from the heat of the sun, carbon-fiber structures surrounding the battery cells conduct heat to the outer surface of the rover. At night, Red Rover will hibernate, and it will awaken with the sun to fire up nonaqueous lithium iron phosphate batteries rigorously tested by then Carnegie Mellon mechanical engineering undergraduate Charles Muñoz. That is the kind of innovation on the cheap that the X PRIZE is meant to inspire. Although Astrobotic stands a good chance of winning the Google Lunar X PRIZE race, it faces steep competition from India and Russia, which are jointly sponsoring a lunar rover, and from China, which is building a rover of its own that will use a radioisotope to stay powered up through the lunar night. If one of these gets to the moon first, the top prize drops to $15 million. COMPETITION whittaker’s team is also expecting strong competition from other X PRIZE participants. Mountain View, Calif.–based Moon Express, with backing from billionaire co-founder Naveen Jain and other wealthy individual investors, may be the best funded of the Google Lunar X PRIZE teams. It entered the fray only in 2010, three years after the contest was announced, so it is lagging behind Astrobotic. But it is overcoming its latecomer disadvantage with a preexisting spacecraft platform developed by NASA. Another contestant is Boulder, Colo.–based Next Giant Leap, headed by former U.S. Air Force pilot-turned-entrepreneur Michael Joyce. Joyce’s company has teamed up with Draper Laboratory (which designed the guidance, navigation and control systems that shepherded the Apollo spacecraft to the moon), a group at the Massachusetts Institute of Technology, and the space systems branch of Sierra Nevada Corporation. It is building a novel “hopping” spacecraft that obviates the need for a separate rover. The craft reignites the thrusters it uses for touchdown to lift off again and travel short distances to areas of interest. The idea seems workable but only if Joyce can raise the necessary funds. The Google Lunar X PRIZE organizers hope that if they build it, the market will come—that developing rovers and getting them on the moon will spur the growth of a new market. Astrobotic, for example, is offering room onboard its spacecraft and rover at the rate of $1.8 million and $2 million per kilogram (2.2 pounds), respectively, plus a $250,000 “integration fee.” For researchers such as University of Maryland physicist Douglas Currie, at least, a guaranteed spot for a fixed price on a commercial mission would be a boon. Currie and his colleagues want to place an array of laserranging retroreflectors on the moon to support measurements that would be 100 times more accurate than can be made with those left by the Apollo astronauts— if only missions become available on which to fly them. Perhaps the most enduring benefit of the X PRIZE will be to inspire the next generation of scientists and engineers. The race has lent an air of real-world excitement to Whittaker’s Advanced Mobile Robot Development course. During the final week of classes in April 2011, members of the Astrobotic structures team scurry about the 3,000-square-foot workshop of Carnegie Mellon’s Planetary Robotics Laboratory, which is entirely dedicated to the moon rover project. They are testing the design for fragmenting metal bolts, an alternative to typical explosive bolts, that unhinge the ramps from the spacecraft so that the rover can explore the lunar surface. Grad student Kanchi Nayaka and a group of undergrads prepare a high-speed video camera on a tripod to record the simulation. The students then throw a switch, and 17.9 seconds later the bolt breaks apart with a bang, and the ramp swings open and falls to the ground, ready for the rover to emerge. “Awesome!” Nayaka says. She steps back from the camera and shoots a grin at a visitor. “You must be good luck!” The most enduring benefit of the Google Lunar X PRIZE may be inspiring the next generation of scientists and engineers.

Scientific American, April 2012