Specifically in the case of complex mechanical space vehicles (e.g. rovers, landers), do we just refuel and send the battle tested and carefully fine tuned prototype on the mission, or do we take the risk of trying to make a near perfect replica and send it after basic lab testing?

Did the actual Curiosity rove Earth before Mars?

  • $\begingroup$ If you are asking about potential cost savings, then there are none: for something like Curiosity, you need an identical vehicle back on Earth anyway, to be able to replicate and study any unforeseen problems in a lab setting. So, you might just as well send the shiny new one. Commercial sattelites are a different matter, though. Here, cost does matter, and we don't expect unexpected things to happen. $\endgroup$ Nov 13, 2014 at 17:51
  • $\begingroup$ Actually my question is about reliability. I assume that building a space vehicle involves a lot of manual assembly/craft/adjustments which may be difficult to exactly replicate on a second vehicle. This is contrary to e.g. car manufacturing where tools and processes are very mature so the output is very consistent. $\endgroup$ Nov 14, 2014 at 6:16
  • 1
    $\begingroup$ To clarify my previous comment:In some rare cases, things may work precisely due to unknown imperfections. The test vehicle may have these imperfections and pass the validation test, while the launch vehicle may not have them and fail on the mission. $\endgroup$ Nov 14, 2014 at 6:23

3 Answers 3


Normally you build 2 prototypes of a spacecraft; one is called the engineering prototype (or model), and the other one is called the protoflight.

Both must be identical, but the engineering model is normally reassembled several times during development and testing, and will stay on earth as a faithful reference of the flight model.

The protoflight must be handled with extreme care (to avoid any damaging of connectors, or ageing of mechanical components). It is the protoflight that must undergo all final qualification tests (thermal-vacuum, vibration, acoustic, outgassing).

Note that for beasts such as telecommunications satellites, there is no budget for building a complete engineering model. These are integrated based on already tested and reliable components (avionics, propulsion, solar panels, transponders, antennas, etc.).

  • $\begingroup$ I spent a whole year working on an engineering prototype satellite. Sometimes the budget IS there for that. We did every possible test on that thing except sending it into space. We did send it into simulated space for two weeks though. It wasn't a test of anything new, the customer just wanted to be sure. $\endgroup$
    – Jasmine
    Nov 14, 2014 at 18:16

Curiosity drove in a clean room on Earth, but it never got dirty until it got to Mars.

Curiosity driving on ramps in clean room

Different engineering models were used to test those aspects of the rover in the dirty outdoor Mars Yard at JPL.

In general the actual flight vehicles are treated somewhat schizophrenically, handled very gently and with great care in clean rooms, then beaten all to hell in vibration and shock testing and frozen and heated in thermal vacuum chambers.

Generally the "battle-tested" engineering model is not "fine tuned". It is hay-wired, jury-rigged, and duct-taped as needed to get it to work. The flight hardware implements changes learned from the engineering model as well as for other reasons, e.g. changes in interfaces, and also it has all flight components where there may have been compromises in the components used in the engineering model. It is a very rare, perhaps mythical, engineering model that can be flown.


In the context of Interplanetary contamination, any craft that leaves earth for an interplanetary mission from most countries falls under the Planetary protection section of the Outer Space treaty.

The COSPAR defines categories within mission targets and indicates the precautions to take.

Mars is in category III/IV, and therefore requires the craft to be perfectly clean. The decontamination procedure is described on the planetary protection page. It includes clean room assembly and components sterilization before assembly. I am pretty sure this rules out the possibility for the protoflight to make a test run outside the clean room before takeoff.

  • 2
    $\begingroup$ Not true. Such Planetary Protection (PP) rules only apply to spacecraft going to potentially life-bearing worlds. No PP applies when going to Mercury for example. $\endgroup$
    – Mark Adler
    Nov 14, 2014 at 15:43
  • $\begingroup$ Apparently I read that a bit too fast, it seems you are right, and it is a bit more complicated. I will edit my answer to reflect that. $\endgroup$
    – njzk2
    Nov 14, 2014 at 15:50

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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