An advanced alien species in my story created smart matter similar to utility fog, nanobots with extendable arms which link to form programmed structures. When the species moves to new locations a small volume of nanobots can be programmed to replicate.

Eventually some of the nanobots no longer respond to their programming and like a grey goo scenario, they begin to self replicate and form structures of their own designs.

As the bots have only two abilities, form structures and replicate, once all raw materials for replication have been used up on the planet, how do they travel to the next celestial body for more materials? Without their programming evolving intelligence to the level that they can build/form their own propulsion systems, how would they travel through space to the next location?

  • $\begingroup$ What type of 'programmed structures' do they normally build? Why? Wouldnt that already include some form of movement/propulsion system that can be redesigned by the new bots? $\endgroup$ May 2 at 20:21
  • $\begingroup$ What do the new bots 'structures of their own designs' do? Why? Why/how did they escape their programming in the first place? Answer these questions, and you'll be half-way to answering your own question. $\endgroup$ May 2 at 20:24
  • $\begingroup$ @WolfieSmith I had not considered mechanical forms as their original use, I envisioned rooms with all solid objects forming but not electrical equipment, not sure why, maybe they would be too complex? Their own designs are random fractal forms, basically they look nice or weird but serve no purpose. currently how they escape programming is a vague fault in programming or software. $\endgroup$ May 2 at 20:35
  • $\begingroup$ When the species moves to new locations a small volume of nanobots can be programmed to replicate... Eventually some of the nanobots no longer respond to their programming and... they begin to self replicate and form structures of their own designs where is this advanced species when this happens? are they still in the planet? they have already left? also, what kind of "raw materials" do the nanobots need? how do they identify, detect, or look for them? $\endgroup$
    – Josh Part
    May 3 at 18:53
  • $\begingroup$ @JoshPart Wiki says they can be made of Aluminium Oxide. Self-replicate is a term used a lot when I read up on utility fog but how they do it I don't know, it seems like more complex features are needed or other machines are needed to manufacture and mine materials. $\endgroup$ May 3 at 21:27

9 Answers 9


Q: Without their programming evolving intelligence to the level that they can build/form their own propulsion systems, how would they travel through space to the next location?

They would not

Without propulsion of some kind, these nanobots won't be able to leave the planet. For that, there is a minimal escape velocity.. without propulsion, your bots could float up to ca 8000 meters using wind and turbulence (upward airflow, soaring) but I don't think they can reach a higher altitude than that.

Also, I'd like to note that "self-replication" does not imply "self-modification". The robotic entity needs additional parts, new mounting points, and probably some kind of docking station ("home")


Two ideas:

Forward thinking.

This requires simple pre-programming to initiate:

When they survey the planet's structure, they'll find the volcanic potential. Stripping the planet's atmosphere can be done in time by absorbing the gasses, converting them into solid form, salts and the like - it'll take energy, but is necessary.

They then seed areas around volcanoes and super volcanos with toughened encapsulated versions of themselves with a large surface-area to volume ratio - this so that when the volcanos erupt, the spores will be ejected into low-orbit, the star's wind will then carry the lucky few out of the influence of that planet, with a chance of entering the atmosphere of another and starting the cycle again.


This requires more complex pre-programming in order to achieve along with heuristics and an idea about the complexities of language:

They listen, they watch. When they detect advanced technology producing radio-waves or other signs of life, they pay attention to the transmissions.

They can then learn to communicate. Strategies might include:

  • Identifying and repeating simple distress calls: "SOS"
  • Asking for help - "we're stranded on ah hostile planet, help us!"
  • Expressing interest in contact - "we're not space faring, but we'd love to meet you. We have fun and interesting things to share with you, secrets of pleasure/immortality..." etc..
  • $\begingroup$ Atmospheric stripping to facilitate volcanic space launch - nice! But I was disappointed to hear the "secrets of pleasure" message is from grey goo because I was going to answer that one. Maybe I still will; you never know. $\endgroup$
    – Willk
    May 2 at 22:27

(Assumes that either they draining resources from smaller asteroids rather than terrestrial planets, or that they eat planets so thoroughly that basically all of the planetary mass is converted to goo, with no atmosphere remaining.)

The Jorō spider is known for being beautiful, for being aggressively invasive, and for flying. When introduced to a new area, they set up their meticulously-crafted webs, decimate the local insect population, and then breed the next generation. Those hatchlings use their silk not for webs, but for "parachutes". They extend a strand into the air where the wind can catch it, and catch it it does, lofting the hatchling into the air. It can't control its direction, but when it touches down again, it will try its best to build a web at the new location, continuing the outward invasion.

The Jorō goo behaves similarly. After devouring a planet, an insanely large number of bots rest in their own gravity well, exposed to the harsh radiation of the sun. Those who feel the hardest of the hard radiation extend a small sail into the solar wind. At first, nothing happens; resources to build a big enough sail to escape a gravity well just aren't to be found. But eventually, within a few decades, the ejecta from a solar flare washes over the planet. The magnetic fields of the CME charge the sail, and the high-energy protons in the ejecta slam into it. For most of the trillions upon trillions of goo bots, nothing happens. For a small percentage, the radiation destroys them, or nearly so; flips bits of memory and corrupts their software and behavior. And for a tiny bit of a tiny bit, enough momentum hits at the right angle to fling the bot out of the gravity well and into space.

With trillions of trillions of bots on a planet, its hard to imagine that only hundreds or thousands may escape the well. But once they do, they travel through space for a long, long time. Some are accelerated by the solar magnetic fields into interstellar space; some end up in an orbit that goes no where. But if even one should land on another planet that can be mined...

The invasion will continue.



Evolution is a game of numbers.

Once the nanites are no longer following pre-programming, they follow evolution: Whatever makes more nanites survives. So if some of your wild nanites retain the memory of ship building, they will travel to new mass sources and replicate more nanites. If they re-evolve the ability to make some kind of propulsion, they may travel to new mass and replicate. Perhaps they make a spore structure that drifts through space until it randomly hits a mass. They could build giant cannons. They could take over existing ships and use them instinctively or follow the stolen ship's programming.

Perhaps they retain a small number of 'tame' nanites pre-differentiated to make engines. These nanites build engines, the other nanites consume. As long as the builders make engines and the consumers retain a breeding population of engine makers, all is good. In this scenario, the tame nanites are like organelles performing a specialized function needed by the rest of the organism.

You might even have stealth nanites that sneak along with nanites that can make engines. Then they out-compete them at the destination. Perhaps these ship nanites then evolve defenses to get rid of these stealth nanites to keep the mass all to themselves.

So there are lots of ways your nanites can get around. But the ones that can't, don't. They eat and stop. If they find a way again, they go. The ones that can, do. They go on.

  • 1
    $\begingroup$ -1 For evolution you would also need nanites to have mistakes in reproduction and (ideally, for any reasonable efficiency) some form of mixing of those mistakes. Basically you need some analogue to genetics. The idea of dynamic specializations is not an automatic given. $\endgroup$ May 3 at 7:18
  • $\begingroup$ @DavidMulder Evolution is not just genetic. If you read my answer, the examples I gave were primarily behavioral. The whole premise of the question is that the nanite replication has a programming mistake that makes the nanites go rogue, and without alteration and mistakes, this would not be possible. So the possibility of mistakes is in the question. $\endgroup$
    – DWKraus
    May 3 at 11:02

Impact Scenarios

Assuming your nanobots are hardy enough to withstand their own weight as they eat the planet, some would survive an impact event and could end up in space that way. Not purposefully traveling, but destined to land somewhere eventually.

Generic Survival

Assuming they are NOT hardy enough to withstand the weight of millions of others above them, perhaps they simply react to spread out. I don't know how feasible the math is, but they may baloon out to a vast honeycomb structure or a single spinning thread instead of a solid planet. They would basically be trying to get away from the center of gravity as much as possible, and perhaps the inverse square law would save them. If they manage to do that, the left over rotation from the planet may spin the outer layers off into space. I'm thinking that even if they all formed a strand one nanobot wide stacked up, they'd be thousands of times longer than the planet's original diameter, but the gravity would technically be the same on the "lower" ones. Even at one bot wide, there's still millions of bots stacked up. But a string that long with even a little spin will probably experience enough inertia (centrifugal force) to pull them off into space. Basically, a space elevator, with no planet left to hold them down.

An actual space elevator

If one of the structures they build when working properly is a space elevator, even just the static non-moving parts, then some nanobots are already in space as the planet changes. Even without the overall mass changing, there's a good chance that the rogue nano-bots may just break down the cable or stop participating in the cable at some point, and everything above that gets flung into space.

  • $\begingroup$ I like the space elevator! Even if not programmed ever to do so, the particles may just by curiosity or error build a space tower on themselves after already multiplying for quite some time.. Using fractal structures, which they already know for building every other large construction, it won't take a large mass of particles to reach space beyond the 'ejection height'. Particles then would easily launch from the tip of the tower, as they where never programmed to work properly beyond gravity. $\endgroup$
    – dronus
    May 3 at 18:50

They will have no ability to travel through space unless their creators first transported some into space for the purpose of utilizing their abilities in engineering projects in zero g conditions outside their own planet's gravity well. If that occurred and some went 'rogue' while deployed in the vacuum of space on for example a construction mission its possible low mass structures could be propelled by the pressure of solar radiation and thus escape the bounds of the solar system.

Problem is though the journey would be more or less in random directions (mostly away from the system elliptic and slow with a capital S. It would tens of thousands of years to pass the closest star to their point of origin and chances of actually entering a star system randomly by chance would be astronomically small. So your probably looking at journeys millions of years long.

Alternate Scenario/frame challenge: Your robots could actually be good candidates for a self repairing solar sail. Low in mass, potentially highly reflective and self repairing provided a suitable supply of raw materials were part of the payload. So perhaps instead they were deliberately launched on an interstellar and after many years, for some reason (radiation damage?) forgot their original mission parameters beyond reaching another system and making more sails.

However while bots that managed to make it and stay in zero g might be able to repeat that process any that fell into all but the smallest gravity wells would be destined to fail. Even if the local environment didn't destroy them all they could do is start mass producing sails on the surface. And any sails they made would in turn be doomed to degrade due to local environmental forces. So they'd more than likely to just end up in a closed loop - constantly recycling deceased bots from old sails into new ones. Large parts of the planet might end up nice and shiny though - for a little while.


If they are in a gravitational gradient, they can do sittups to increase their height: (chapter 2) https://assets.researchsquare.com/files/rs-451074/v3_covered.pdf?c=1631871991


Think Katamari Damacy. You gain power by gaining mass, you use that power to gain more mass, etc. Start small, converting thumbtacks and bouncy balls, and work your way up to planets and stars.

All known means of propulsion in space require ejected matter of some sort. If you're a giant cloud of grey goo, the only matter you have available to you is - grey goo.

If your goal isn't to get the entire cloud into space, but merely a minimum viable reproducing subset of the cloud, you would propel yourself by ejecting large numbers of nanobots in the direction opposite to your intended travel.

The range of travel is limited by the size of the cloud, but if the cloud is able to find matter (asteroids, moons, planets, and yes, maybe even stars), and convert them into more grey goo, it can acquire more mass and thus more nanobots to use as propellent in order to continue travelling.

If a grey goo cloud consumes an asteroid and turns it into grey goo, it has that much more matter (in the form of grey goo) to expel and propel itself to other sources of matter. If it consumes a moon, or a planet, it has even more. By the time it grows enough to consume entire stars, its ability to expand would be practically limitless.


They Don't

The inability to leave the planet is an intentional safety feature. Since the nanomachines son can only form solid structures, the grey goo can only ever eat a single planet, rather than the whole solar system.


You must log in to answer this question.

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