Autonomy

Which technologies provide the best return-on-investment for intelligent rovers?

Whenever the Mars Pathfinder rover experienced a failure, it had to stop, wait for the next scheduled opportunity to communicate its problem to Earth (relatively brief periods each day, due to limitations of the rover's solar batteries), and wait for new commands attempting to resolve the problem. After each command, the Earthbound controllers would await Pathfinder's progress report before issuing a follow-up command.

The process, guided by extreme caution, was tedious and time-consuming. The twin MERs (Mars Explorer Rovers) currently on Mars follow a similar procedure. Technology that would increase a rover's autonomy -- that is, improve its ability to conduct science while reducing its need to phone home for help -- would save a great deal of time and therefore enable the rover to accomplish much more.

These studies represent efforts to determine the relative benefits of investing in various software technologies that purport to help Mars rovers do science more efficiently, avoid most failures, and diagnose and correct their own problems when failures occur.

The first study (Rover Autonomy Study #1) focuses on technologies that were proposed specifically to reduce fault rates observed during extensive field-testing in Mars-like terrain here on Earth.

The second study (Rover Autonomy Study #2) analyzes technologies that were funded as basic research, only loosely coupled to a mission. Hence, we needed to determine technology-derived capabilities and match those capabilities with mission requirements. These technologies are more advanced than those in Rover Autonomy Study #1, capable of automating entire sequential operations.