Fail-proof is not possible. Fail-tolerance is the goal. Engineers design for robustness, and try to include enough redundancy in critical systems to allow work-arounds in the event of failure. It is too costly to provide back-ups for every system on a probe, and some systems are too heavy and bulky to have back-ups, so these systems must be robust. Data pathways, chipsets, etc. are generally lighter and smaller and can be backed up. The design of any probe has to revolve around costs, payload weight, fuel requirements, etc, etc. If you design in too much redundancy, the weight of the probe increases, the amount fuel required to launch it increases, the amount fuel required to maneuver the probe increases. All these increase project costs and reduce the number of "nice to have" and "optional" sensors, experiments, etc that your friends and associates want to tack onto the probe.rhia said:
rhia said:
