So how do you maximize the scientific findings you get from a robot if you can't even move it from point A to point B without years of effort and the risk of irreparable damage?
Put a drone on it.
Engineers at NASA JPL (seen in the above video) are working on a proposed add-on to future rovers, such as Mars 2020. In theory, Mars drones would be great: They could soar over the rough surface of the red planet, quickly imaging large areas and selecting the best touch-down spots to optimize scientific analyses. Even if the helicopter was separate from the science-doing rover (as the current prototypes are) they would still be able to pick a perfect route and destination for their partnered robot, maximizing the rover's success.
Unfortunately it's not as simple as popping a commercial drone into space. Mars has less gravity, but it also has a much thinner atmosphere. Since copters get their lift by cutting through dense air, a copter on Mars needs to be designed to compensate. Researchers also need to make sure that the rover can always land perfectly when it touches down. A rover doesn't need to crash and burn in order to be dead on arrival: A poor landing can mean the robot is just slightly askew, leaving it unable to right itself or expose its solar panels to enough light.
Every day, scientists watching on earth would have seven seconds of terror as they waited for their drone to touch down safely. But with the potential to triple the daily distances currently surveyed by traditional rovers, the risk could well be worth it.