First, CubeSats were invented, those breadbox-sized probes packed with microelectronics and sensors that allowed university labs and small businesses to explore space. A number of CubeSats have already been launched from the International Space Station, and plans are afoot to send some to cis-lunar space and in pursuit of Earth-approaching asteroids. Now, Astrobotic and Carnegie Mellon University are developing the CubeRover under contract from NASA.

The idea is that groups of two-kilogram rovers would land on the moon or Mars, either along with a big rover or in place of it, and then fan out across the landscape, covering more ground and going to places where it might not be safe to send a larger, more expensive vehicle.

The CubeRover would cost a fraction of the amount of a more conventional rover, which can range from a few hundred million dollars to billions. Just as with CubeSats, CubeRovers can be developed by small labs and businesses and included as secondary payloads on space launches.

Thus far the development team have managed to create a rover as light as six kilograms and have a design for a four kilogram module. By contrast, the car-sized Mars Curiosity masses just short of 900 kilograms.

The way that a mission to the moon would work using CubeRovers would involve landing as many as ten, twenty, or more at a single site. Then the CubeRovers move out in different directions to perform experiments and prospect for minerals and other useful materials.

Either the CubeRovers will have their internal power sources, say solar cells, or they will be powered by batteries and would return to the main lander from time to time for a recharge.

Another version of the mission would include a conventional rover and a swarm of CubeRovers. In this scenario the CubeRovers would travel as a kind of advance recon force for the main rover, scouting out interesting routes, noting what hazards may be in store.

The main rover could serve as the recharging station for its tiny siblings, In theory, this robotic scouting party could cover a great deal of ground during the length of a mission, even if a certain percentage of the CubeRovers come to grief in the interim.

The CubeRover development project is a bargain for NASA, costing only $125,000 for the first six months and then a few hundred thousand more as things progress.

NASA has awarded Astrobotic a number of these small business contracts.

Astrobotic aspires to be a shipping company to the moon, delivering payloads for both government and private customers. The company was, at one time, part of the Google Lunar XPrize competition but choose to drop out in favor of more long-term development work. It has plans to do a first lunar landing in 2019.