NASA scientists and engineers are testing new technologies using a K9 rover in a newly built ‘Marscape’ test facility in preparation for future missions to Mars.
Testing is being conducted at NASA Ames Research Center in California’s Silicon Valley in a 3/4-acre ‘Marscape’ that has been designed to resemble the terrain on Mars. Constructed at a cost of about $74,000, the test facility incorporates the environmental and geological features of Mars that hold the greatest scientific interest. The Marscape features a dry lakebed and outflow channel, a meteorite impact crater, a volcanic zone containing a dry hydrothermal spring and an area that scientists describe as “chaotic terrain.”
“The goal of the K9 project is to integrate and demonstrate new robotic technologies that will enable NASA to meet the science goals of future Mars missions,” said Maria Bualat, a computer engineer at NASA Ames who is the K9 rover project lead. Scientists hope to utilize new robotic technologies during NASA’s Mars Science Laboratory (MSL) mission anticipated in 2009.
“The whole purpose of this research project is to ensure that this rover is as autonomous and reliable as possible. Autonomous instrument placement capability is essential for future Mars exploration,” said Dr. Liam Pedersen, principle investigator for the K9 rover instrument placement project.
Developed jointly at NASA Ames and NASA’s Jet Propulsion Laboratory (JPL), Pasadena, Calif., the K9 rover is a six-wheeled, solar-powered rover weighing 145 pounds (65 kg) that measures 63 inches (1.6 m) high. The K9 rover is modeled after a rover named “ FIDO” (Field Integrated Design and Operations) developed at JPL about four years ago.
The rover’s avionics, instrumentation, and its autonomy software were developed at NASA Ames. The rover carries a variety of instruments on board, including a compass, an inertial measurement unit and three pairs of monochromatic cameras used for navigation and instrument placement. The rover also carries a pair of high-resolution, color stereo cameras and the CHAMP, an arm-mounted, focusable microscopic camera developed at the University of Colorado, Boulder. The rover’s stereo cameras create a 3-D virtual map of the exploration site that scientists use to help navigate the rover to its intended target.
“Approaching science targets such as rocks and placing instruments against them to take measurements is an essential task for a planetary surface exploration rover,” Pedersen explained. “This is necessary to acquire samples, determine mineralogy, obtain microscopic images and other operations needed to understand the planet’s geology and search for evidence of past or present life.”
Due to Mars’ distance from Earth, even with commands being transmitted at the speed of light, it currently takes three martian days to complete the process of directing a rover to a targeted rock and placing the instrument on the rock to begin scientific analysis of it. Scientists at NASA Ames hope to be able to accomplish that objective in a single day, thereby increasing the efficiency of obtaining science data in future missions.
David Smith, a computer scientist at NASA Ames, leads the research group that is responsible for developing the rover’s automated planning and scheduling software. In previous missions, there has been very little automation of the planning and scheduling process for planetary rovers, according to Smith.
“What’s unique about this software being developed at NASA Ames is that it generates contingency plans to provide an alternative that can be executed when things go wrong,” Smith said. “There is a great deal of uncertainty in operating a robotic system on Mars, so you need to be able to consider alternatives. By having options available, you increase the science return.”
To increase the versatility of the software, scientists at NASA Ames, JPL and Carnegie Mellon University are developing a universal architecture for robotics software named CLARAty, funded by the Mars Technology Program, to develop robotics capabilities at NASA centers and universities for future missions.
“NASA near-term Mars missions have very ambitious science goals that will require high levels of autonomy onboard the robot,” said Bualat. “Our goal is to have a ‘smart robot’ that we can send off to Mars in 2009 that will take care of itself.”
The K9 rover project’s annual cost of approximately $1 million is funded jointly by the Intelligent Systems Project under the Computing, Information and Communications Technology (CICT) Program administered by NASA’s Office of Aerospace Technology, and by the Mars Technology Program, administered by the Office of Space Science, NASA Headquarters, Washington.
Reproduction quality images of the K9 rover are available at: