For a rover to demonstrate robust autonomous operations, it will need to adapt to its environment. Although someday this may be performed through automatic, on-board modification of plans and models, this could produce unanticipated behaviors and thus would introduce an added risk into the mission. Instead, we focus on those aspects that can allow the rover controllers on ground to specify the rover's response to a range of possible operating situations that may arise. In particular, we consider the following criteria essential for robust rover autonomy:

• Robust flexible operation
  A rover should choose its behavior based on the execution environment. The rover operators should be able to specify multiple execution behaviors along with the conditions under which they are appropriate.

• Resource utilization
  Resources such as power, data storage, and communication bandwidth place restrictions on what the rover can do. In order to perform its science and mission objectives efficiently, the rover needs to make full use of the available resources, even when they change from expected values. To do this it should track actual and expected resources, signal resource conflicts, and allow its behavior to be specified with respect to the resources.

• Failure recovery
  Because of the complexity of a rover's interactions with its environment, it will need the ability to recover from faults or anomalous situations. The rover should recover from those faults from which it safely can, and it should gather relevant information for problems that need to be handled by rover operators. Each of these will save command cycles and increase the efficiency of the rover.
  In addition to these criteria, there are other considerations that affect planetary rover operations in general. These considerations influence the design of missions and consequently the work described in this paper.

• Uncertainty
  Rovers operate in environments that are largely unknown. Techniques that require precise information about the environment to perform actions will be prone to failure.

• Need for safety
  Space missions are expensive and rover failure irreparable. A planetary rover must work the first time, and it needs to guarantee its own safety in order to achieve the mission goals.

• Limited communication
  With limited use of the Deep Space Network and with other communication limits imposed by orbiter schedulers, the rover must operate with relatively little and infrequent interaction with the rover operators. The rover must operate between communication events with no human intervention and no assistance from ground-based programs.
• Need for understandability
  Mission designers must be convinced that the proposed approach will improve the quality of the mission. If the approach is complex and unclear to the mission designer, it is unlikely to be used.

• Multiple objectives
  A deployed planetary rover must balance considerations such as navigation with science, communication, resource consumption, and fault recovery. A rover must be able to operate as an integrated system.