A lot of the costs of launching space probes to other moons and planets must be the staggering research and engineering costs to design the vehicle as well as the costs of planning a mission.

In addition, there's often limited windows of time required as the planets all need to be aligned in a certain way.

So with that in mind, why only launch 1 or 2 probes at a time? Why not launch 100 (obviously, just a figure off the top of my head) all at once? Obviously, the cost of launching 100 space probes will be more than 1, but surely, it's not going to be 100x more expensive. You've already paid for the fixed costs - you just have to pay the variable costs like the fuel and parts for each ship.

Introducing redundancy also minimises risks of the mission failing. These missions often require some pinpoint accuracy and a failure at any stage may end the mission.

  • 2
    $\begingroup$ "Why not launch 100 ... all at once?" If they're going to the same place by the same path, there is the risk of collision, so why not join them all together in the first place (and be able to share the common systems)? $\endgroup$ Jun 16, 2015 at 3:16
  • 4
    $\begingroup$ "Introducing redundancy also minimises risks of the mission failing." Not if they are launched in a 'batch' and the launcher blows up. Multiple Hubble Space Telescopes would (presumably) all have been out of focus, and all of the Mars lander vehicles that crashed due to a misunderstanding between metric and imperial measurements would have met the same fate.. $\endgroup$ Jun 16, 2015 at 3:20
  • 2
    $\begingroup$ Batch production would lower development cost per unit and the Mars2020 rover will indeed be a Curiosity copy with different instruments. But with a budget of almost 3 billion, launch cost of maybe 0.2 billion is not too important. Also, scientists want to send different and unique things to get broader data (like: flyby, orbiter, lander). The launcher is more suited for batch production and I think it is done for Ariane 5. $\endgroup$
    – LocalFluff
    Jun 16, 2015 at 4:40
  • 5
    $\begingroup$ "Obviously, the cost of launching 100 space probes will be more than 1, but surely, it's not going to be 100x more expensive. " The fallacy here is the implicit assumption that costs won't rise linearly, but somehow benefits will. If the cost of 100 probes is only 2x the cost of one probe, it still doesn't make sense if the expected benefit is only 1.5x that of a single probe. $\endgroup$
    – NPSF3000
    Jun 16, 2015 at 12:01
  • 2
    $\begingroup$ Other factors are the need for facilities for collecting the probes' data and sending commands to them: is there enough bandwidth in the system to handle 100 probes simultaneously? Then you need people to plan the probe's mission (which for a Cassini or Curiousity is non-trivial), scientists to analyze the resulting data, etc. $\endgroup$
    – jamesqf
    Jun 16, 2015 at 18:31

6 Answers 6


The cost savings from mass production are not linear. You will see savings in some areas, but none in others.

  • Space probes, and even commercial satellites (which are standardized to an extent) are hand-built. So if you build 2 (or 20) the labor cost rises more or less linearly. Only when you build much larger numbers (thousands), it becomes interesting to invest in production tooling to exchange for lower labor cost.

  • The same goes for the components you'll use. 100 probes isn't close to the scale you need to profit from mass production.

  • Capability does not rise linearly with size. If you have a choice between 1 large probe and 100 small probes, the large probe can contain much more advanced science instruments than 100 small ones. Outer planet missions need an RTG, and these are rare enough that building 100 of them isn't feasible at the moment. A small probe won't be able to transmit directly from e.g. Mars to Earth, it doesn't have the transmitter power budget.

  • There are cases where you'd want to have multiple measurements at different locations. Conveniently, an orbiting spacecraft will visit every location on the planet eventually (depending on the chosen orbit). NASA is planning several missions that will contain a few cubesats in addition to the main spacecraft, to do measurements in several places at once. The main craft will act as a relay.

  • Deep space launches are hard. For outer planet missions, the current probes are near the launch weight limit of the rockets we have. So sending 100 Voyagers would have required 100 launches. Sending 100 spacecraft in 1 launch would have meant building cubesats - something that's only recently become feasible at all.

I guess my thinking is that if a company spent all their resources designing an innovative new product, they're not just going to want to create one. They're going to want to create many - because each unit they produce will lower the overall costs per unit.

For now, deep space probes are generally designed and built by space agencies and/or universities, not commercial companies. They don't care about mass production.

  • $\begingroup$ I think the final point ("Deep space launches are hard") is the major killer of this idea. Keeping in mind that you decide on the launcher first, then build as much science into its payload capability as you can for the desired trajectory and mission lifespan. There won't be a lot of cases where a large number of small probes on an identical (or nearly identical) trajectory will deliver more science than a single, larger probe of the same total weight, although obviously some exceptions to that rule can clearly be found. $\endgroup$
    – user
    Jun 16, 2015 at 15:27
  • $\begingroup$ Note though that there are a few instances of multiple, basically identical probes. The Spirit & Opportunity Mars rovers and the Voyagers come to mind. But AFAIK the largest number of identical probes launched at nearly the same time is two. $\endgroup$
    – jamesqf
    Jun 17, 2015 at 6:06
  • $\begingroup$ Just to clarify, the only orbit that "visit[s] every location on the planet eventually" is a polar orbit. Other orbits can cover a lot of area, but only a polar orbit will cover everything. Also, I think if your orbital period is a multiple of the rotational period of the planet you will wind up covering the same ground repeatedly, instead of covering everything. Still, as you say, most orbits cover a significant area! For instance, the ISS does not pass over every point on earth, just a large band centered on the equator. $\endgroup$
    – Dan
    Aug 4, 2015 at 17:23

A batch of probes all going to the same destination wouldn't be economic at the moment. Each probe would need its own propulsion, communications (unless there's one "parent" with a long range antenna), guidance and other electronics. This would add weight.

Also for general observations, a single probe can just take all of the measurements/pictures and send them back.

However some groups have been looking at a swarm of micro probes. There might be a use for flying them to their destination on one lander then releasing them on the surface for exploration, for example. Another example that's been given is for multiple entry probes to examine Jupiter's atmosphere...


I think the general principle you're thinking about is only partially valid. Here is how I summarize it:

Mass production cannot happen without a factory. Factories can cost billions. But we do reuse the design of probes, especially the early probes in our history.

Research, design, and development have costs all their own. So I think there's a valid part of your principle that says, why design a new probe? We could just make multiple copies and send them everywhere they need to go.

Well, this is true for early space probes. A probe designed to flyby Jupiter would gather the same data from Saturn, Uranus, and Neptune.

And this is indeed how it was done at first. Voyager 1 & 2 were virtually identical, and I'm pretty sure the same is true of Pioneer 10 & 11 and the Mariner Probes. In addition, a lot of the really early probes to the Moon, Mars, and Venus were copies of each other. The Ranger series, for example, goes out to at least Ranger 8.

That does not mean we mass produce them. 8 Ranger probes is no justification for spending billions on a factory when they can just be assembled by hand.

But there comes a point when the same probe just won't gather any new data, because we've already done it! Time to design a new probe that gathers new data.

You may also be wondering about worlds unvisited, or recently visited, using entirely new probes. Why build New Horizons to go to Pluto when we can just reuse the Voyager design, for example? Why spend money on new development (and risk new failures) when we already have a valid design proven to work? The problem here is that the launch budget wasn't allocated for a 720 kg probe (Voyager's mass). Instead, we wanted a 480 kg probe (New Horizons' mass) so we could hurl it out to Pluto's distance really fast. Also, these days, designing a new space probe is NOT as hard as it used to be (thankfully).


Besides the arguments about the efficiencies, probes are specially-designed objects with lots of discussion about the particular payloads.

With a serial-launch scenario, you get to incorporate your understanding of the object from early probes into the design of the later probes. With a mass-launch scenario, you have to decide on all the payloads with your current knowledge of the target. It would be like playing "20 questions", but you have to submit all your questions at the beginning of the game rather than develop them over time.


According to rocket equation, you will need fuel in proportion to your payload mass and exponentially in relation with mass exhaust velocity. Indeed, maintenance costs will be less in batch launches, however, you will have to use more fuel which has its own consequences on launch preparation. In addition, today's launchers are not designed to be chained in arbitrary configurations, i.e, you cannot stack five upper stages over thirty core stages.

  • 1
    $\begingroup$ It would be very nice if we could stack rocket launchers in arbitrary configurations. $\endgroup$ Jun 16, 2015 at 7:38
  • 1
    $\begingroup$ "According to rocket equation, you will need fuel in proportion to your payload mass squared." Why is that? $\endgroup$
    – JiK
    Jun 16, 2015 at 9:51
  • $\begingroup$ @JiK: Infinitestimal stage approximation. $\endgroup$ Jun 16, 2015 at 10:08

There are definite cost savings when launching multiple satellites. A company I am familiar with launched many communication satellites. Using a common design saved engineering costs, some labor (Things got faster with time, and less down time for experts), software, and even testing (Some satellites were more heavily tested, reducing the cost). You are absolutely right, more probes would be less expensive. However, there are some drawbacks.

  1. The amount of science gained from a new probe goes down with the number of spacecraft. For instance, MRO has a camera that has photographed the entire surface of Mars at fairly high resolution. Right now it can only take pictures of the surface to see things changing. It's practical science value has dropped dramatically.
  2. Communication with the group of satellites becomes difficult. One of the most difficult aspects of managing many space missions is when they are all in the same part of the sky at the same time. This reduces the availability of communications windows, making it harder overall to hear from these probes. So many at one location would be very difficult.
  3. So many missions at once would greatly reduce the effective budget of NASA, making it difficult to launch new missions in the future. This would end up with less science, as new better instruments are available all the time.

Making a duplicate of an instrument can be useful. For instance, the two Voyager probes are clones, as were the two Viking landers. This level of redundancy can be helpful, and doesn't cost significantly more.

You might be interested to know there is a mission in the planning that sounds very much like yours. It is the expedition satellites planned by Planetary Resources to identify satellites to mine for minerals. It will take essentially a duplicate satellite and send it to many different potential targets. This makes perfect sense, to use low cost copies of probes to study many targets around the Solar System, learning as much as one can.

  • $\begingroup$ Excellent response. I'm fairly certain that MRO has not imaged all of Mars at its highest resolution. I don't think there is enough bandwidth (even over years) to send that data back. $\endgroup$
    – Freddo411
    Dec 18, 2021 at 23:42
  • 1
    $\begingroup$ @Freddo411 It has at its second highest resolution, using Context. With HiRISE, it's actually around 10% of the planet imaged right now, but the bandwidth is a huge constraint. $\endgroup$
    – PearsonArtPhoto
    Dec 19, 2021 at 14:23

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.