OK, so the main problem with Beagle 2 may have been inadequate funds (surprise!) or poor risk management but a pertinent factor had to have been the control of the a tricky process like parachute landing through a planetary atmosphere.
Near open-loop control is viable when the craft under control is in a predictable medium (such as vacuum) and where the craft is essentially a rigid body. The many successful journeys to the gas giants exemplify this approach. With proper flight planning, there will be plenty of time for corrections (maybe even with an Earthman in the loop) if a manouevre doesn't get the craft into its target state.
This could be the killer application for spacecraft autonomy (more so than, say, scheduling of obervation modes). Big guns are working on it.
I would guess that some of this will be spin-off from UAV tech (very hot at Farnborough this year).
Others are more excited at the thought of autonomy in the process of going up, rather than down.