With a 2.7s time delay, rover operators on Earth cannot exert real time control.  They have to upload tasks, which the lander and rover must work together to accomplish.  Source: Control Engineering

The rules state: “To win the Google Lunar X Prize, a team must successfully land a privately funded craft on the lunar surface and survive long enough to complete the mission goals of roaming about the lunar surface for at least 500 meters and documenting it with a video program, called a “mooncast,” sent back to Earth. “The mooncast consists of digital data that must be collected and transmitted to the Earth composed of the following:

• High resolution 360º panoramic photographs from the surface of the moon;
• Self portraits of the rover taken on the surface of the moon;
• Near-real time videos showing the craft’s journey along the lunar surface;
• High definition (HD) video;
• Transmission of a cached set of data, loaded on the craft before launch (e.g., first e-mail from the moon).

“Teams will be required to send a mooncast detailing their arrival on the lunar surface, and a second mooncast that provides imagery and video of their journey roaming the lunar surface. All told, the mooncasts will represent approximately a Gigabyte of stunning content returned to the Earth.”

There are, at the time of this writing, 10 teams officially registered in the competition. Team Frednet is a multi-national team comprised of systems, software, and hardware developers who serve as the leaders and overall coordinators of an international group of open-source developers, engineers, and scientists. Their goal is to bring the same successful approach used in developing major software systems to bear on the problems associated with space exploration and research.

The controls problem the team faces once they get to the moon is how to steer their rover. It takes approximately 1.35 s for a radio signal to span the 385,000 km (239,228 mi) average distance between the Earth and the moon.

“What we’re imagining,” says Dan Smith, hardware engineering lead for Team Frednet, “is a very small rover with dual systems for communication and video, and the capability to orient itself and have some fault tolerance and fault correction built into it.”

The system is more than just one spacecraft. The team envisions three separate components: a lunar orbiter (called the “lunar bus”), a lander, and a rover. These components will start out like Russian matryoshka nesting dolls, with the rover carried by the lander, which will, in turn, be carried by the bus.

The whole assemblage will be lofted into low Earth orbit by an as-yet-undetermined commercial satellite-launching service. From Earth orbit, the bus will carry the project to lunar orbit, where it will deploy the lander. The bus will remain in lunar orbit to help relay signals to and from the team’s base on Earth.

The lander will descend to the lunar surface and deploy the rover. As the sole stationary component, the lander will include radio systems to communicate with the lunar bus (still in orbit), the rover, and directly with Earth.

The rover’s job is to move about on the lunar surface, posing for cameras based on the lander and acquiring imagery of its own. The lander will transmit the rest of the required mooncast data.

The lander will carry preprogrammed commands and procedures. From Earth, team members can send commands to initiate programs and procedures. With the long communication delay, however, they cannot exert real-time control. That capability must be shared between the lander and rover.

“The rover wants to have general autonomy for figuring out how to get from point A to point B, aim its antennas, aim its camera, and send signals back,” says Smith, “The lander is actually supposed to act kind of like a server — watching out for the rover like a mother duck.”