Self replicating robots - how realistic?

[Sky Captain]

In Sci-Fi it is common to have robots that you can dump on a planet/moon/asteorid and task tem to build more of themselves and then to build whatever you want from local resources. I'm wondering how realistic such machines are?
For example in real life to build a complex machine you need parts and materials made by many different factories that use raw materials mined often in different continents with help of dedicated heavy duty mining infrastructure.
Now consider a planet with no infrastructure. You drop few of your robots there and order them to replicate. How are they going to mine all required raw materials with no mining industry? How are they going to locate all neccessary ores? Some metals needed for advanced electronics are in very low concentrations and you have to process tons of ore to get few grams - task that pretty much require heavy machinery. Some components require advanced manufacturing processes typically done in huge factory.

Are there any realistic means how our current mining and manufacturing industry could be miniaturized enough to fit into something that can be easily carried aboard spacecraft deployed anywhere you want?

[Stark]

It depends if you want them to replicate, or replicate efficiently for a designed purpose in space. 'self replicating robots' are generally conflated with implacable AI and super lasers.

[Starglider]

Is this with or without nanotechnology? Conventional industrial processes exploit economies of scale by preparing large batches of resources and components in isolation, and then progressively assembling them into the final product. Nearly all our manufacturing processes are optimised for bulk processing, extraction works by scooping up lumps of stuff and then removing all the bits we don't want, even a lot of manufacturing consists of removing material from an initial blank to make a desired shape. Computer chips for example take a big building full of infrastructure even if you are using e-beam litho to eliminate the need to make masks.

Living organisms by contrast do all their assembly in parallel with the materials and components intermixed in a water solution. They extract millions of different compounds from the same input material using sophisticated filtering. Living organisms make little use of rare elements, partly because the ability to structure at the molecular scale removes the need. Proposed nanotechnology assembler designs share many of the attributes of biology, in particular that they do not rely on economies of scale for acceptable performance. For this application we'd be talking about dry nanotech architectures that aren't constrained to water solutions and protein chemistry. This makes a huge difference in performance and feasibility; not just the assembly part, directly nanoassembled computers for example would be far less reliant on rare elements. You don't need the full-blown and debatably plausible general free-motile nanoassemblers for this, Avatar quite sensibly went for 'stereolithography plants' that are advanced versions of current prototype 3D printers.

[Imperial528]

With modern knowledge, I believe that making a single machine capable of making more of itself from scratch is pretty hard to do, and likely not at all in the present day a feasible concept. Even if one was made that worked, I doubt it would be capable of handling the more complex tasks required for things such as a Von Neuman machine.

However, we can do the next best thing. Send out a system of robots, a few mining types, refining types, fabrication types, and assembly types, and if needed a separate control type. If designed with modular construction and standardized parts in mind, the entire system could replicate itself and then build a new rocket in probably a few years at least.
Sure, it would require a few decades of R&D to make it work right, but the technology and scientific knowledge is there, even if the methods to fully utilize it are not yet realized. There are places working on simmilar concepts throughout the world, and one that comes to mind is these cube-shaped robots that can assemble themselves into different configurations without any outside instruction, and they can actually learn to assemble more efficiently. I forget who made them, though.

[Sky Captain]

When asking about physical possibility, I like to point out that such machines do exist today: plants and animals.

Well, the ''infrastructure'' to support complex life took millions of years to develop. In this case we would want something along the lines of dump few machines on desert planet, come back few years later and have everything from housing to spaceships built and ready for use.

Is this with or without nanotechnology?

Nanotechnology is allowed however it has to be kept realistic in a way that we should be reasonably sure that it would work once the engineering difficulties are sorted out.

[Dr Roberts]

When asking about physical possibility, I like to point out that such machines do exist today: plants and animals.

Well, the ''infrastructure'' to support complex life took millions of years to develop. In this case we would want something along the lines of dump few machines on desert planet, come back few years later and have everything from housing to spaceships built and ready for use.

Yes but you are forgetting that was because it was "chance" that created life but this would have an intelligent designer.

[lordofchange13]

When asking about physical possibility, I like to point out that such machines do exist today: plants and animals.

Well, the ''infrastructure'' to support complex life took millions of years to develop. In this case we would want something along the lines of dump few machines on desert planet, come back few years later and have everything from housing to spaceships built and ready for use.

Yes but you are forgetting that was because it was "chance" that created life but this would have an intelligent designer.

Being artificial or not is irreverent it would still take time for the machines to convert the atmosphere in to somthing earth like.

Then these super magic robots have to find(assuming that they can't transmute sand in to uranium, or iron/gold/copper/boron) would have to find large material deposits of raw materials, then fabricate some sort of refinery as well as a smelter, and a factory to make other kinds of robots. But the machines still need power for all this manufacturing so they have to make either a fission reactor or find oil well(unlikely sense the plane did not support life before) to a generator. The OP states that all of this has to be made out of modern technology and be able to be light enough for a existing rocket to get it in space all the way to another planet. in short we have no technology that could achieve what you ask in a reasonable amount of time.

[Spectre_nz]

Well, the ''infrastructure'' to support complex life took millions of years to develop.

super magic robots

We've already got those too...
Microbes are hardy little fuckers.

There are Clostridia that given a smattering of elements like iron, nickle, cobalt, zinc, selenium, tungsten, calcium, sodium and potassium (they rather like chloride salts) will happily build everything they need out of Carbon monoxide, hydrogen sulphide and ammonia. Supplementing it with some vitamins makes it go a lot faster, but it's got the gene pathways to live without them.
Basically, there are microbes out there that don't need anything more than what comes out of a deep sea volcanic vent.
I'm pretty sure we'd find a lot of archaea that can perform similar tricks if we looked in the right niches.

Tinkering with some microbe to get it to do what we want is a lot easier than designing tiny robots from the ground up.
For the terraforming part anyway. I wouldn't trust their workmanship to build me a star-port...

[lordofchange13]

Well, the ''infrastructure'' to support complex life took millions of years to develop.

super magic robots

We've already got those too...
Microbes are hardy little fuckers.

There are Clostridia that given a smattering of elements like iron, nickle, cobalt, zinc, selenium, tungsten, calcium, sodium and potassium (they rather like chloride salts) will happily build everything they need out of Carbon monoxide, hydrogen sulphide and ammonia. Supplementing it with some vitamins makes it go a lot faster, but it's got the gene pathways to live without them.
Basically, there are microbes out there that don't need anything more than what comes out of a deep sea volcanic vent.
I'm pretty sure we'd find a lot of archaea that can perform similar tricks if we looked in the right niches.

Tinkering with some microbe to get it to do what we want is a lot easier than designing tiny robots from the ground up.
For the terraforming part anyway. I wouldn't trust their workmanship to build me a star-port...

But their is still the problem of terraforming a whole planet in resemble amounts of time.

[Spectre_nz]

Any terraforming attempt is going to take a significant amount of time unless you resort to magi-tech.
As the OP was asking for things we could realistically do with our current technology, terraforming is going to take an exceptionally large amount of time assuming we can even get there. (Granted, he was asking about mining, but microbes can help facilitate that)

Microbes have a giant advantage in that they've had billions of years to get very good at the exact things we want here; self replication and persistence in hostile conditions. They'll survive in space, you can freeze dry them, fire them across the cosmos, culture a larger population when you get to your target, then you seed the entire target globe with a range of species who'll scratch out a niche on the most basic precursors.
You can pulverize the crust from orbit then send in the microbes. They'll modify your target's atmosphere, leach metals out of the pulverized rock to speed your later collection and engage in bio-warfare with any hostile native life, if that's what you're into. Granted, that's easier said that done, but bacteria already do all these things, so we just have to modify, rather than build from scratch.

Then in a hundred thousand years when you've sent in a second wave of yeast and algae to make a breathable atmosphere you can drop prefabs on the surface and mine the old fashioned way. Tedious and boring, but it has a proven track record...

[lordofchange13]

Not contesting that, just saying that a human government/organization would not invest for a pay off several hundred thousand years away.

[Sky Captain]

Microbes are good at self replication and survival in hostile environments, but how feasible it would be to genetically engineer a strain of microbes that for example can "grow" metal parts for some sort of machine and then somehow assamble those parts into final product?

Proposed nanotechnology assembler designs share many of the attributes of biology, in particular that they do not rely on economies of scale for acceptable performance.

How fast would be the construction process? It seems that if you are assambling something literally atom by atom it would be very slow to make large scale structures.

[Rabid]

I don't know how realistic this proposal is, but in his story "The Nanotech Network", the author proposed to circumvent the difficulty of building objects atom-by-atom by using micro-scale cells that were an hybridization of a living cell and inorganic nano-scale material.

In this proposal, the cells themselves where used as the basis of everything built, like some sort of electronic-clay. This allowed for and incredible flexibility and speed when it was needed to reconfigure an object for a different function.

In the scope of this thread, it is interesting to see, in the link produced above, that this concept is being extensively worked on right now by major corporations ; albeit sans living cells in it for the moment.

On the matter of self-replication, I don't know if those hypothetical "claytronic" nanobots will have the capability of reproducing themselves without intermediary, but they could at least I think build some sort of mini-factories to build more of themselves : thoses factories just won't be at the nano-scale.

[Dr Roberts]

Guys you have to remember that idea behind it is to send thousands or even millions of the machines and terraforming can be pretty easy depending on the atmosphere.

[Spectre_nz]

Getting them to spit out metals in an elemental form isn't something I've heard of them doing, I have a hunch the reducing equivalents required are just a bit too high.

However, they will 'smelt' (ie, bioleach) several metal ores quite happily (albeit, slowly) leaving you with a concentrated liquor of metal ions you can pull out of solution with electroplating.

That may not sound impressive, but it reduces large sections of the infrastructure you need to storage tanks and your ezy-grow microbial culture. Shit, you could imagine automated crawlers that chew up rock and digest it in a microbial loaded stomach vat before passing the liquor on to the electroplating intestine and it all starts sounding a lot like you've got a biological organism than a factory. Those metal leachers need oxygen, which means you'd need to keep them indoors on some inhospitable world, making atmosphere for them with something like a blue-green algae. Basically you'd need to get your own little ecology going, admittedly, that isn't easy.
If you're going to do it with anything, however, microbes are the easiest place to start.

Building the big stuff would be beyond their skill set, but if you could marry them with a 3D printer you could start churning out components.

[Spectre_nz]

yeah, wow, I left that txt window open for friggin ages and didn't check new replied.

Guys you have to remember that idea behind it is to send thousands or even millions of the machines and terraforming can be pretty easy depending on the atmosphere.

So, if you've got the capacity to send thousands or millions of machines to terraform a planet in record time, you've got magi-tech. Or something close to it.

The point of self replication and other what not is so you don't have to waste resources lifting a million machines, you just lift a few, then make the rest there. Then turn them into low cost housing afterwards. Then you can spend your savings on millions of machines moving colonists, trade goods and other handfuls of machines to other planets.

[Dr Roberts]

The concept is still to send loads. The point is each one is pretty cheap (relatively). I was watching Physics of the impossible presented by Michio Kaku where his idea was to send these off in to the galaxy where when they find a habitable planet they would construct a base where you would then beam the info of how to clone a body and download their mind in to the new body. This would of course be done under extreme mental conditioning for them to accept they are clones though.

[Spectre_nz]

The concept is still to send loads. The point is each one is pretty cheap (relatively). I was watching Physics of the impossible presented by Michio Kaku where his idea was to send these off in to the galaxy where when they find a habitable planet they would construct a base where you would then beam the info of how to clone a body and download their mind in to the new body. This would of course be done under extreme mental conditioning for them to accept they are clones though.

That's not exactly a unique concept. The point is, how would you actually go about doing this in reality, and what have we actually got that would allow you to do some of these things.
And in reality, send more mass and launch it into space don't exactly see eye to eye.

The concept is to send enough to get the job done. The less you need to send to get the job done in a time you deem reasonable, the better, as that weight saving means you can send other useful equipment, or equipment to other locations so all your eggs aren't in one basket. Brute forcing the problem has its merits, but you still want efficiency. You've sent a million tons of robots to terraform a planet? Why not also make those robots self replicating, and then leave it to them to make their own way to the next planet and get to work there so you don't have to make and launch another million robots next time.
Launching a million robots to a million planets beats launching a million robots to one planet. (assuming it works...)

Why send loads when you can send a few and they'll make loads on site?

Moving mass from one star system to another using plausible methods, those well beyond what we can do now but not physics breaking, is expensive, due to the fuel requirements, and the requirement to carry more fuel just to carry the fuel that you're carrying...
Anything you can do to reduce the total mass you need to boost out of your star's gravity well means you save resources to spend elsewhere.

[Dr Roberts]

The concept is still to send loads. The point is each one is pretty cheap (relatively). I was watching Physics of the impossible presented by Michio Kaku where his idea was to send these off in to the galaxy where when they find a habitable planet they would construct a base where you would then beam the info of how to clone a body and download their mind in to the new body. This would of course be done under extreme mental conditioning for them to accept they are clones though.

That's not exactly a unique concept. The point is, how would you actually go about doing this in reality, and what have we actually got that would allow you to do some of these things.
And in reality, send more mass and launch it into space don't exactly see eye to eye.

The concept is to send enough to get the job done. The less you need to send to get the job done in a time you deem reasonable, the better, as that weight saving means you can send other useful equipment, or equipment to other locations so all your eggs aren't in one
basket. Brute forcing the problem has its merits, but you still want efficiency. You've sent a million tons of robots to terraform a planet? Why not also make those robots self replicating, and then leave it to them to make their own way to the next planet and get to work there so you don't have to make and launch another million robots next time.
Launching a million robots to a million planets beats launching a million robots to one planet. (assuming it works...)

Why send loads when you can send a few and they'll make loads on site?

Moving mass from one star system to another using plausible methods, those well beyond what we can do now but not physics breaking, is expensive, due to the fuel requirements,

and the requirement to carry more fuel just to carry the fuel that you're carrying...
Anything you can do to reduce the total mass you need to boost out of your star's gravity well means you save resources to spend elsewhere.

Check out sci fi science physics of the impossible how to colonise the galaxy.

[Spectre_nz]

Oh do fuck off and take 'I saw it on TV, lol' with you.

Rather than reference a video clip you saw one time and not even providing a link, do you have anything to say on why the efficiency of moving less mass isn't useful and why brute forcing the problem is somehow better?

Or have you just failed to get the point I was making entirly?

We know you can just send more and do it faster, dumbshit, but if you could do that in the first place, then you weren't utilizing your full transport potential. If you are maxed out on the potential mass you can transport, you can always get more bang from your buck by also making what you're sending self replicating, if you can pull that trick off.

[Dr Roberts]

Dude your the one with the fucking inability to comprehend such a simple fucking concept. The point is that even self replicating robots aren't worth it if you only send a few. The galaxy is pretty damn big. If you think sending a thousand of even self replicating robots is efficient then you are a dumbass. That is why you send millions of self replicating robots (I DIDN'T SAY SENDING NON REPLICATING ONES!) If you are going to try to make someone look stupid make sure you are disagreeing with them first.

[Spectre_nz]

Aaaaaaand do you know what you could have done from the outset?

Explained your points so you didn't sound like such a retard. Turning around and saying its everyone else's fault because they can't magically distil an essay out of your paragraph is pretty damn weak.
Quoting my entire post and saying 'go watch a video' is lazy in the extreme.
That's my main beef here. You running in like you're still in show and tell.

If you'd started out with something along the lines of "Self replication is certainly desirable, but you're still going to need redundancy, so you'd have to send many..."
Then you'd be doing better. But no. You breeze in and drop "O Hai guize, you kno you send heaps rite? Terraforming's easy if you find a nice planet."
Its like you're a) Not actually paying attention to what's going on in the thread and b) missing the obvious.

No one has actually talked about how many you'd send or the size and mass of each individual. Sending as many as you can affordably transport is numbingly obvious, but 'As many as we can' depends on what it is you're sending.
And while that video touches on what we're talking about, it goes into very little depth and doesn't really address anything beyond what's already been said in this thread. Sure, it throws in a few sound-bites, suggests you could build a self replicating nano-swarm, but doesn't go into any detail as to how the fuck you'd do that. Cut to, the OP, here, asking how the fuck you'd actually do that.

Does you knowledge of robotics, nano-technology, mineral recovery or space travel extend beyond what you sponged out of that video? Are you able to build on the discussion so far? Or did you just want to show everyone a cool videoclip you found and run off?

[Number Theoretic]

About ten days ago, Slashdot linked a report about a recent success in developing self-replicating molecules. They use DNA for that, but according to the article, they use it in such an entirely new way that it is safe to say, these molecules are artificial or synthethic.

Judging from that, i'd say in response to the OP: yes, in principle, this might be possible. However, this does not say anything about on which scale, how fast and in which environments. I believe that is still relatively open and some of the more constructive posters here have already written about possible ideas.

[NoXion]

Regarding atmospheric alteration using genetically engineered microbes; surely if one designs a microbe that can thrive in the initial environment, then one would be able to take advantage of the power of exponential growth? Sure, there may be other factors in terraforming a world that would slow down the process, and one might need to re-inoculate the atmosphere with a different strain to keep it balanced right, but considering the importance of marine algae in maintaining a breathable atmosphere, I shouldn't think establishing one would take too long if the required elements for it were mostly already on the planet.

[Sky Captain]

That may not sound impressive, but it reduces large sections of the infrastructure you need to storage tanks and your ezy-grow microbial culture. Shit, you could imagine automated crawlers that chew up rock and digest it in a microbial loaded stomach vat before passing the liquor on to the electroplating intestine and it all starts sounding a lot like you've got a biological organism than a factory. Those metal leachers need oxygen, which means you'd need to keep them indoors on some inhospitable world, making atmosphere for them with something like a blue-green algae. Basically you'd need to get your own little ecology going, admittedly, that isn't easy.
If you're going to do it with anything, however, microbes are the easiest place to start.

Building the big stuff would be beyond their skill set, but if you could marry them with a 3D printer you could start churning out components.

Yeah, that sounds like something that could be based on technology although still beyond our current level, but at least understandable to us, maybe even possible to build at the end of this century, not your typical semi magical replicator nanobot swarm.
One could imagine a ship carrying several of such machines enter orbit around planet then scan the surface to determine the mineral composition and then deploys those machines in a place where there is the most favorable mineral composition.
There probably would be several mining robots that dig up various rocks then robotic trucks carry raw materials to refinery that with help of genetically engineered microbes leach out various metals then refine them and feed to 3D printer that can produce all kinds of componenets and then feed those componenets to assembly robots that build final product. Advanced AI would be required that runs the whole process and can solve all kinds of problems that would be encountered.
Still sounds like a fairly large system that probably would mass thousands of tons, but still that is better than carrying your whole industrial base with you together with an army of engineers and technicians.