Juno got me thinking again about scaling and maximizing the output of space exploration missions. It seems weird that (with the notable exception of Mars rovers) we (as in humanity) always only start each probe once and then more or less throw away the design. Given the large up-front investment (allegedly 1 billion $ for Juno) and the long latency between start and scientific results, why do we not invest a little more (I assume roughly 10%) to launch a second or third probe of the same design? After all, production and launch cannot be as expensive as design and control. By this reasonign we could visit all big outer planets for less than double the cost.

The advantages would be clear: A second probe could provide redundancy in the case of an unexpected problem (mixing metric/imperial units, radiation, micrometeorites, etc.) a third probe could be launched to a different target (i.e. Saturn) and while possibly not being ideally designed still yield interesting results and given the different travel times and launch windows make better use of earth-born control infrastructure.

Is this really just a funding thing (i.e. the budget being as tight as possible) or is there a more fundamental reason behind this?

edit: Just to clarify, I am not talking about mass-production. I am talking about sharing the common cost between a few (maybe up to a dozen) variants of the same probe-design. My question is: How much could be saved when a design is reused and why it seems to be not that much.

  • $\begingroup$ "..to launch a second or third probe of the same design? .. By this reasonign we could visit all big outer planets for less than double the cost." Probes for different planets would have different heating/cooling requirements, different amounts of radiation protection, antennas designed for that distance... As such, it would make sense to tailor the probe to the expected environment. So no, that logic is way off, also I think your "(I assume roughly 10%)" estimate is, like most stats, made up on the spur of the moment (as well as being wildly optimistic). $\endgroup$ Jul 5, 2016 at 8:51
  • 2
    $\begingroup$ @MichaelKjörling: There's no Acme Spacecrafts Ltd, but what about Jebediah Kerman's Junkyard? $\endgroup$
    – Tim Čas
    Jul 5, 2016 at 11:42
  • 1
    $\begingroup$ @TimČas Feel free to pay a visit to my garage. Third on the right after the Great Red Spot. You might find something useful. $\endgroup$
    – user
    Jul 5, 2016 at 11:45
  • 2
    $\begingroup$ Interestingly enough, we actually did this with the Viking Program and landed two practically identical landers on Mars. You'd have to dig through the cost breakdowns to see what kind of percentage went to the overall design, but I'd imagine it's not as high as you think and launch/construction is still the bulk of the cost. $\endgroup$
    – M.A.H.
    Jul 5, 2016 at 14:15
  • 1
    $\begingroup$ In the early days of space exploration, this was often the case; most probes were launched in sets, in case one failed. For example, the Mariner missions sent to Mars were paired; Mariner 3 failed, but Mariner 4 succeeded. However, in those days, the chances of failure were extremely high; nowadays, we know what we are doing, and our failure rate is very low. $\endgroup$
    – Phiteros
    Jul 5, 2016 at 17:21

1 Answer 1


We definitely don't throw the design away. We improve upon it, keeping the good, improving the weak, discarding the bad and adding the new and promising.

Launch is expensive. Control is expensive and must be done on per-item basis, not just per-design. Construction isn't exactly peanuts either.

The cost estimation for scientific and experimental craft is:

enter image description here enter image description here

abbreviations used:

  • TT&C telemetry tracking and command
  • ADCS attitude determination and control subsystem
  • EPS electrical power subsystem
  • IA&T integration, assembly, and test
  • SE/PM systems engineering/program management

You'd save up much on SE/PM and some on IA&T. The rest, for two probes is almost doubled (discounts for doubling the purchases or streamlining construction may or may not apply), and that even doesn't include launch cost (pure doubled). Mission upkeep cost would be reduced, but then... we'd be getting mostly duplicates of the science. Missions are tailored per destination; it makes no sense to send the exact copy of a probe meant for Jupiter to Venus, so the benefits wouldn't be doubled; they'd be increased but not nearly by as much as you seem to think.

So, unfortunately, your guess of 10% is way off. The figure of 60-80% is much more likely, and the returns wouldn't entirely justify the expense.

And with the budget available, unfortunately, "less than doubling the cost" means in fact "almost halving the budget per craft". It doesn't mean we're getting twice the science per mission; it means we're getting half the science per probe! Splitting the budget will mean less instruments, less delta-V, less everything per craft.

So - definitely not worth it.

There is an option where that makes sense though: a failed mission. RUD on launchpad, probe critical failure, not reaching the destination. Of course new improvements and fixes to detected problems would apply, but building a very similar probe for re-launch totally makes sense... if you can find money for it.

  • $\begingroup$ Thanks for the in-depths answer. If these estimates apply to Juno that means NASA pays roughly 100Mio dollar for 60m² of solar panels, 110Ah Lithium batteries and necessary wiring and control. Even worse, they are paying roughly 80Mio dollar for the structure alone. That's the price of a LEO launch. Wow. $\endgroup$
    – choeger
    Jul 5, 2016 at 9:23
  • 1
    $\begingroup$ @choeger: That's not entirely so; these charts are for craft alone, not for the mission as a whole, also scientific instruments may vary wildly in price, some sensors may be $100, some are $100mln. Never mind propulsion, which will cost an order of magnitude more for that sort of mission. Check the document linked for comparison of Galileo vs smaller crafts. And for Juno, working on solar panels, not RTG, at Jupiter distance I really wouldn't be surprised if the cost of the power system was this much. $\endgroup$
    – SF.
    Jul 5, 2016 at 9:29
  • 1
    $\begingroup$ oh, also, take prices of components from the time the craft was built, not from the time of arrival. The price of solar panels went quite a bit down since the launch. And it's not just panels. Batteries that are to sustain the probe in Jupiter shadow, years after manufacture date, control and regulation systems that must work in the radiation of Jupiter, redundancy for everything that matters and isn't too heavy to be doubled... $\endgroup$
    – SF.
    Jul 5, 2016 at 9:32
  • $\begingroup$ If its only the craft, then what fraction is that of the overall mission cost? $\endgroup$
    – choeger
    Jul 5, 2016 at 11:00
  • $\begingroup$ @choeger: I'd need to dig deeper into the sources and unfortunately I don't quite have the time currently - maybe someone else will answer, or if you have the patience - a lot can be found in that paper I linked. $\endgroup$
    – SF.
    Jul 5, 2016 at 11:22

Not the answer you're looking for? Browse other questions tagged or ask your own question.