For the purposes of this question, assume that a Dyson Array is a collection of discrete solar panels or mirrors orbiting the sun, all of whom share the same orbit and have same velocity.

  1. Can a civilization similar to us with the same level of technology as present day humanity build a Dyson Array for energy supply, say by using asteroids as raw material by crushing them?

  2. If not, what would be the minimum amount of technological advancement necessary to be able to build that?

Please note that the question is not about a solid ring, solid sphere or a spherical swarm.

  • $\begingroup$ The answer or (1) is obviously "no" or we'd do it. It's a major undertaking for us "just" to send a probe to look at solar system objects. Seems a pointless question here. $\endgroup$ Dec 10, 2018 at 12:04
  • 1
    $\begingroup$ @StephenG we could build mirror+boiling oil farms in deserts and dams to store energy for time periods when the deserts dont create enough and power the world with it. We dont because we use kapitalism and dont have a unified earth yet. Just because we dont do something smart doesnt mean we cant. $\endgroup$
    – Demigan
    Dec 10, 2018 at 12:09
  • 2
    $\begingroup$ @Demigan I don't think you understand the scale of a Dyson ring. If we could construct such a thing any problem we currently have in managing global resources would be trivial to solve by comparison. $\endgroup$ Dec 10, 2018 at 12:26
  • 1
    $\begingroup$ @StephenG I understand that the scale and necessary requirements are massive, my point is that "No or we'd do it" isnt a good argument. Another energy production method that is well within our grasp and doesnt require world-scale production and organization isnt used either, doesnt mean we cant. $\endgroup$
    – Demigan
    Dec 10, 2018 at 15:02
  • 1
    $\begingroup$ @StephenG: The obstacles to building such a Dyson array aren't technical, though, but economic & social. That is, we wouldn't have to invent any new science to do it. $\endgroup$
    – jamesqf
    Dec 11, 2018 at 19:16

3 Answers 3


For the purposes of this question, assume that a Dyson Array is a collection of discrete solar panels or mirrors orbiting the sun, all of whom share the same orbit and have same velocity.

In practice you probably don't need to be quite so precise about orbits and velocities. You just need to keep them apart and not blocking each other.

Can a civilization similar to us with the same level of technology as present day humanity build a Dyson Array for energy supply, say by using asteroids as raw material by crushing them?

No even close.

We have none of the required tech except in the basic form for some of the least difficult parts.

If not, what would be the minimum amount of technological advancement necessary to be able to build that?

A rough list of things you need to have are :

Self-replicating space craft.

You can't possibly build and launch enough craft on Earth to build such an array/ring. You need to build machines that you can send out into the solar system that will find and collect resources (probably mine asteroids) and be able to construct copies of themselves and send material to where it's needed to build an array of solar collectors. You also need machines that build the machines that build the solar arrays. The chemical science requirements are mind boggling.

And these machines themselves need lots of energy so just finding a source for that energy (more solar collectors probably) would it's own challenge.

We can't do any of these things now, of course. We're hundreds of years away from this, and I mean just the mechanical aspects.


Doing the above implies advances in chemical engineering beyond anything we can achieve now. It involves a machine finding an asteroid, analyzing it's contents and being able to break it up and refine those into required materials. Some of the materials will be quite rare.

Think about the size of a modern chemical plant and then imagine somehow making something even more complex in space from raw materials.

After you gather it and refine it, you then need to make things. These will include computer components and other unknown technologies we don't yet have possibly. So imagine microchip manufacturing being done on this scale in space as part of a vast production system spanning the solar system with millions of machines working cooperatively but without human intervention (or very little, anyway).

The chemical science to do these things is way beyond us. Again hundreds of years perhaps.

Physics and engineering.

Strangely we probably have enough physics to do these things, but there is a world of difference between physics and engineering.

We probably need some advances in propulsion systems but they're something we could probably get to reasonably easily based on current ideas and prototypes.

Engineering is making stuff and maintaining it. That's a whole other ball game. We have essentially no experience of automated engineering in micro-gravity and the conditions in "raw" space. The challenges are huge. Just designing devices that function for extended periods (we could need hundreds of years to build this array) and do so reliably enough to self-repair or at least recycle themselves is daunting. Again we're a long, long way from this kind of goal. We're still finding new ways that machines can fail after years in space. Finding solutions to those problems is equally hard.

AI and computing.

You possibly don't need AI for this, but by the time make the systems you'll probably have AI and need it to control the systems you build. You may even need the AI to do the research for you to enable this, because there is an argument that doing this kind of advanced science is not practical for humans with limited lifespans and relatively slow learning skills. So maybe we need the AI just to do our thinking for us ! By the time you're building a Dyson array you may already to using AI to do a lot of your thinking (and hence control) for you. AI or something close to it may be a requirement for this. And building an AI isn't trivial and although you hear the expression bounced around a lot now, these devices are only prototypes in basic data processing compared to a proper AI that can "think" for itself. That's another thing that is a long way off.

Even without AI you certainly need a lot of processing power and advances in algorithm design and programming to enable engineering on this scale. Bigger and more complex systems. We've already hit practical limits that mean we're moving away from "faster and bigger" to "more efficient, slower and more parallelism", so there are significant issues in all of that. So budget some time and resources for that too.

What for ? Using the energy !

Dyson arrays sound cool, but there are two problems with this idea that are pretty fundamental :

  • How do we transmit the power we gather ?
  • What do we do with the power anyway ?

This is kind of important. Maybe a K2 civilization has some useful purpose for all this power, but it's just useless to us. We have no idea what could usefully be done with it. Oh, there are wild ideas out there (e.g. build your own black hole), but the reality is that we have no more idea how to use this amount of power than Humphrey Davy would have about using a nuclear power station's output.

So before you start rearranging the entire solar system, you need a purpose !

Transmitting the power is pretty darn important. You have to find a way to move the vast power output from the array to where you can use it. Or we're all moving to the array ? Or something.

And an array like this is also something else : a weapon of vast scale. Anything that could e.g. send vast amounts of power over solar system distances would be absolutely deadly in any hands. Who precisely would have charge of this ? Or we don't transmit the power, in which case we have to use it in situ. which requires building at least an industrial system based around these arrays.

This also raises a problem with that vast swarm of self-replicating semi-autonomous machines. They're potentially just as dangerous in the wrong hands (or some AI's hands !). They'd be designed for micro-gravity, so there's no danger of those machines landing on Earth, but that doesn't mean they (or someone/something controlling them) can't build different machines that can ! And by the time you reach this level of technology you will already have significant numbers of people ling in space - possibly even most people, because self-replicating machines can build vast space stations as well and we'd have cracked 100% recycling by that time.

So we have now a vast and potentially deadly array of machines controlled largely by other machines that can build other machines rocketing around the solar system. Up hands all those who didn't see any Terminator movies and feels safe about this ?

So there are some issues with building a Dyson anything that are only cool if you're a Bond villain.

  • $\begingroup$ Minor nitpick: I doubt we'll ever crack 100% recycling. Really, really, really close? Sure! But 100% is pretty darn near impossible. Certainly not a "next few hundred years" kinda thing. Other than that, great post and +1. $\endgroup$
    – Gryphon
    Dec 30, 2018 at 18:33
  • $\begingroup$ I tend to agree on most parts, but the chemistry to do all that is not difficult. Engineering autonomous chemical plants will be tricky, but the computers and more complex parts you can fabricate on earth, at least to begin with, or in human-controlled factories in space. $\endgroup$
    – Karl
    Jan 17, 2019 at 16:22

In theory, we are already capable of doing something like this. We can send Satellites in an solar orbit, we are able to send energy over long distances via microwave or high-powered laser, and we are most likely able to mine asteroids for its material.

But as StephenG stated in the comments, we do not do this currently.


Because it is not viable, too expensive and possibly even dangerous.
First of all, we would need a stable orbit around the sun. If we would get slightly lower than the earth's orbit, we would not really have some sort of Dyson Ring, but it would work. Interestingly enough, it is much harder to send stuff into a lower orbit, than into a higher orbit, which is the reason why we sometimes send stuff to Jupiter, just to use the gravity assist to slow the spacecraft down. Scott Manley has a great video about that, which I just cannot find right now.
Additionally, it is way too costly to launch stuff into that orbit. You could launch those satellites in GEO and have almost the same effect.

Second, we do have the technology to send energy over long distances, but in space this is a bit harder. Microwave senders lose focus over long distances, thus losing potential energy. Lasers lose their focus at some distance as well. And if you miss your receiver with a high energy laser beam, you can imagine what happens.

Third, we are not really ready to mine asteroids. Again, it is a question of cost/investment. Is it really worth to do that compared to the money you have to invest? You would need drones to do the job, as working in space would have enoumous high health risks for any human. So you need sophisticated software, AI and sensors for that.

Fourth, our current solar panels can get damaged by radiaton, so even if you send up satellites for that job, you have to replace them eventually. And the closer you get to the sun, the more damage they take over time.

What we would need to make it financially viable would be just a few years and a little step in solar technologies. The first company to mine an asteroid, another way to transmit energy savely and of course the software development for AI to do the drone job on their own.

Just to make it clear, this would be just the beginning of building a 'real' dyson ring. We cannot start with millions of satellites, but this would be the basic technologies required for it.

Edit: Just wanted to put it out here, Kurzgesagt has just released a great video explaining what needs to be done and how it could be done.

  • $\begingroup$ Not to mention solar prominences, which could just break apart and melt the dyson array $\endgroup$
    – user55267
    Dec 10, 2018 at 12:54
  • $\begingroup$ @Jannis depending on how close you get, yes. $\endgroup$
    – PSquall
    Dec 10, 2018 at 13:08

We are a whole species, of course, it's possible!

There is a Kurzgesagt – In a Nutshell video about it. They explain science with normal words instead of scientifically ones and they make awesome and cute animated videos about it.

How to Build a Dyson Sphere - The Ultimate Megastructure is one of them. In the 2m 14s, they talk about the usage of millions of solar panels to build it.

I will write here basically what they said.

If each solar panel would have a size of 1km2, we would need around 30,000,000,000,000,000 solar panels and the equivalent 100,000,000,000,000,000,000 tons materials. Obviously, we can't launch that mass into space, that requires a lot of energy and would disassemble an important chunk of Earth.

So, in order to build a Dyson Swarm, we need to cannibalize planets in a cheap way. Asteroids are very small for this, and also they are only largely found in the asteroid belt, which is quite far. Luckily for us, Mercury is a good candidate:

  • It has enough mass to build the Dyson Swarm (we only need to disassemble the 30% of its mass... not a big deal).
  • Has almost a third or Earth gravity (3.7 m/s2).
  • It has no atmosphere (0.5 nPa).
  • Is really close to the Sun.

The last 3 reasons make extremely easy and energy cheap to launch matter from Mercury to the Sun.

So we know our candidate: Mercury.

First, we need to make a sustainable base on Mercury before start making our Dyson Swarm. The base will have primarily robots and only a few humans for very important tasks, with enough money it will be possible to make. The first base there should be in the polar region, which has a constant temperature of -93 °C (maintaining warm the base isn't very difficult). The other option is in the rest of the planet, which temperatures vary from -173 °C to 427 °C (maintaining there is more difficult).

After building the base we can start making the Dyson Swarm.

Building solar panels are very expensive and require constant maintenance, so it isn't useful for our megastructure. As the video says, the simpler the better, so a Swarm of mirrors pointing to the actual solar panels is better.

The first thing to build the swarm would create some solar panels in the Mercury orbit, say 1km2. The gathered energy of these solar panels could be transfer to the planet surface by energy beams, the lack of atmosphere makes this alternative possible.

Once you have a steady supply of energy, the base is ready to start manufacturing more mirrors and solar panels.

  • Launching rockets isn't a good option, they are very expensive, slow and can break.
  • Making a space elevator is really difficult.
  • But taking into account that mirrors and solar panels are autonomous (they don't have people inside), we don't need to take care of acceleration. Building a massive railgun we can launch our probes using electromagnetic forces as if there were bullets towards space.

The mirrors will be sent to the sun while the solar panels will stay in Mercury orbit. Redirecting light from the Sun to the panel, we can gather this energy in the orbit and then transfer it to the surface.

Using exponential grow this becomes quite easy: the first panels produce energy to build the second panel (2). Now both panels produce energy to build other two (4). That 4 to build 4 (8). Now 8 to build 8 (8), and so on. After 60 iterations the Dyson Swarm is ready.

If building the first square kilometre takes a month and the planetary base is able to support the exponential growth, in 10 years it will be ready. Obviously, one thing is the energy grow in solar panels and another the grow of planetary infrastructure, so, in the worst case, we could say that in 100 or 200 years it could be largely done.

Also, remember that you don't need to wait 200 years in order to get the benefit. Even a small part of a Dyson Swarm, like the 0.01%, is able to produce 38,28 Zettawats, enough to melth Earth. You don't need to wait for the total construction of the mega-structure in order to take advantage of its energy.

So, I would say, that given the necessary money and human will, it's possible to actually star building one, at least in an small scale.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .