Robot Colony - Robots prepping for a human mission.

Titan · 2003-06-03 08:14:11

From the way it looks many here on the forums are advocating either nuclear, solar, or wind power generation. If solar is chosen it would likely take a good deal of labor to put enough panels up to generate the necessary power. One option is to make robots do all the work.

A while ago I ran across an article in Discover about robots building robots and sustaining themselves. The energy comes from solar panels and the raw materials come from the soil. The colony would initially start from a small population that is self-sustainable and would reproduce until a target population is hit and the colony would then be switched from reproduction mode to production mode. The robot colony would build supplies for a human mission. I suggest checking out the article in Discover here: Robot, Build Thyself. Here is a link to the abstract from their article in Mathematical and Computer Modeling: Exponential Growth of Large Self-Reproducing Machine Systems.

Palomar · 2003-06-03 09:29:20

*Cool! Thanks for referring this to us.

--Cindy

Josh Cryer · 2003-08-08 12:13:42

Ah, how'd I miss this post? I totally agree with this concept. I think it's rather obvious, given how hard it is for humans to do normal activities in a suit.

I think robots could build anything we needed, really. Cindy was just mentioning pipes. We'd probably need to bury them to keep them from freezing during the harsh winters. You are not going to see a person doing this by hand, probably not even with a big machine like a bulldozer, either. It's a pain in the ass to make controls which can handle harsh environments (which are easy to operate from within a suit), it'd be much easier to have machines which are remotely controlled via some network. If we had a GPS system, or some sort of local surveying equipment, we could plop in a few coordinates, and tell the machine to dig.

Those are some good reads. I hope if anyone comes across other articles of this nature, they would reply here. This is an important aspect of Mars (and indeed, space) colonization.

Josh Cryer · 2003-08-12 11:27:15

I got the paper (Exponential Growth of Large Self-Reproducing Machine Systems), if anyone wants to read it, PM me.

Algol · 2003-08-12 23:04:46

Ive always found this concept really facinating, and i would love to be able to somehow bring it a little bit closer to reality.

Nasa did a study too, im not sure if you have this link.

Advanced Automation for Space Missions

The biggest problem i can see is reducing the size of the initial seed. During the 80's, it couldnt be reduced below 100 tonnes for the initial seed, mainly because you need foundries and smelters for production and these are all heavy items. However over the last 10 years Rapid Prototyping Devices (born out of stereo lithography) means complex machines using a variety of materials can litterally be made in one go from the ground up.

An american company has made the same point at this website.

Ennex

On another note I would love to contine my studies in either of these two areas - if anyone has any ideas, let me know!

Josh Cryer · 2003-08-12 23:50:05

I think the major problem has always been creating a closed model; taking care of the chemistry necessary for little robots to run around using generic processes to sort out whatever they need.

I think that this paper I'm sending you outlines the necessary tasks really well, and pretty much proves what I'd been talking about all along (not sure if you've been following my wonderful magical posts about robot building robots). This is the sort of thing I wish I was paid to study.

The fab technology is good, but auxons are better in that they use the fab technology to make more of themselves to make more fab technology and so on. The human element would then be very minimised. There would be no necessity to go out on the lunar surface with a bulldozer in a space suit. You'd just tell an auxon to go build a certain building using schemetics, and it'll do it.

The size of a seed in my opinion, should be no larger than a small building. A space shuttle payload should be more than enough to get a seed going anywhere. I think smelting processes, though, are pretty much the level we're going to be working on. Most useful stuff in space needs large materials. Being able to build small things is good (and obviously necessary if we want to expand our logic systems and so on- what I'd suggest be done is an auxon be built by the seed specifically relating to ICs and other small things), but we can't use techniques which fabricate small things to build big things, efficiently. So I suspect the first seed would be ?better? at making big things, but that seed could build something which is ?better? at making small things. Specialization is key, because it abstracts tasks to be easier to handle (and that ultimately is what the article talks about).

Josh Cryer · 2003-08-15 15:37:59

I just realized (well, last night, but I keep losing connections here) that auxon technology would be easier to deploy in space. On the moon or Mars (or venus for that matter- floating auxon-plant-hybrids!).

Try explaining a field of robots in Death Valley to the California government. Try to prevent fringe thinkers from going completely nuts and blowing them up (like the Earth Liberation Front).

Auxons wouldn't know how to protect themselves (and rightly so), so we'd have to keep people from going nuts over the technology ourselves.

In space, however, there is no law, no rule. Send a seed to Luna, communicate to it via very high encryption, and you could build whole cities without the intervation of one body; if anything this self-growing city would compel people to get their asses together and go!

space_psibrain · 2003-08-15 18:52:10

Of course the problem on Mars would be communications, unless you had a crew orbiting mars, or using Quantum Entanglement communications to control the bots

Josh Cryer · 2003-08-15 23:02:57

Nah, you don't control every aspect of auxon behavior, you just tell them how they should evolve.

On Mars, you'd just tell them to build something X by Y by Z using certain schemetics, and it'd do it, at the end of the build the last remaining auxons would either be recycled, or reprogrammed to do other tasks which they're physcically capable of (harvesting auxons could go on to recognize garbage so when our crew arrive, they could run about cleaning suspected garbage- just be careful where you put stuff!).

Any mutations would be science studies.

pootechie · 2003-08-15 23:42:44

I think it's rather obvious, given how hard it is for humans to do normal activities in a suit.

Now, I'm a little ignorant of current cutting edge robotics, but I was under the impression that we're a long way from building machines which possess the dexterity and adaptability of even a heavily encumbered human being.

This discussion is quite fascinating, and I feel bad for interupting it even slightly, but I think it should be kept in mind that by the time we're capable of the autonomous alchemy which you folks are talking about, we should be able to do quite a lot of things, even building a better space suit.

space_psibrain · 2003-08-16 01:48:59

Well, the problem with simply putting down humans in suits is more extensive than it seems, because of all the support infrastructure that they would need. So perhaps robotics would be easier...in a decade or so we'll have made very great strides, and I believe there is technology now that could be implemented

Josh Cryer · 2003-08-16 03:01:13

No, no, you're mostly right. But, I have always felt that the chemical part of this sort of technology was the hardest. ?How can I go and dig up so-and-so amount of materials without having to get some materials from the outside to help me??

The paper we were talking about outlines the chemical processes necessary for such a system. Here's a rip from the paper (the paper itself is semi-hard to understand mid-way through; you do need some relative understanding of calculus to wade through it): Click here

A caption for the image (I didn't/won't inline the image because it's rather large and downscaling any more than it is would ruin it:

Figure 2. The element separation cycle. Processing begins with raw dirt as an input at the top of the figure, and proceeds counterclockwise. The products are shown at the outside. With each step is listed the amount of heat which must be added to bring the reactant up to the process temperature and to accomplish the reaction. The heats are normalized to one kilogram of raw dirt. The electrolysis step requires electric energy which is listed separately. Ignoring recovery of heat from the final products and steps with a net energy release, the total energy consumption of the cycle is 25 MJ/kg or 7 kWh/kg.

Basically, though, once we have the energy to do this process, we can pretty much do anything. All that's left is some nifty AI and engineering that doesn't exist yet.

(A seed would probably deploy solar panels, or run off of a nuclear ractor at first; the former probably meaning that initially it would run very very slowly, perhaps as low as a few kilograms of material a day until more solar panels could be produced).

Now, I'm not saying it would be trivial, not at all. But I do think it's within our current capablities. Especially if we do it by breaking down the tasks as much as possible.

We don't need a robot that has a sophisticated hand with the ablity to move in many different ways to bolt a screw; no, all we need is a little crawler robot that can sense a bolt and can know whether or not the bolt needs to be tightened and take the appropriate action (the ?bolt,? of course, could be designed specifically for the robot).

Say we need to build struts for a building, all we'd do is create the struts in the furnace (the furnace would have a nifty mold capable of forming most basic geometric shapes; with an extrude ablity so that shapes that wouldn't fit entirely inside could still be created). Lay the struts out, and our little robots could crawl over them, bolting where necessary.

All that's really necessary, in my opinion, is to design a seed and the inital bots. We'd need some sort of formal proof showing how the machine could create itself (sort of like how programming languages have to have the ablity to be self-hosting to be ?complete?).

If this technology existed, as the paper points out and as common sense would dictate; it would change humanity drastically. Living in space would be a piece of cake, and it would cost nearly nothing.

Josh Cryer · 2003-08-19 04:00:48

I have edited my post to contain a link to the auxon cycle diagram.

space_psibrain · 2003-09-08 23:21:19

Danke

Josh Cryer · 2003-11-19 05:37:18

Here's a link to some papers that talk about the feasiblity of self-reproducable robots, some of the papers talk specifically about lunar robots and how transportation and so on can be easily facilitated.

Source: http://custer.me.jhu.edu/publication/se … ating.html

I think that the rail gun concept is rather common sense, if you want to get X object to Y place on the moon, simply send it into a semi-orbit that has it land there. It's like throwing a baseball, if you could throw hard enough, you could make it land anywhere in the world (although atmospheric friction may become a problem).

In any case, I think that this idea is pretty nifty. And I wish that design considerations were explored more deeply. Instead of prototyping on lego mindstorms, why not create a simulation of a set of geometric shapes that could best form any given metal object? The criteria are still rather large, but it looks like the feasiblity of this idea getting better and better. What we have is the potential to build any given space based object with very little energy, and only time to waste.

If only President Carter's idea for a moon base was realized...

dicktice · 2003-11-19 18:22:16

Josh: Wow! How did that report get by me. Probably while I lived in Sweden, where they were so anti-American anything as a result of the Vietnamn War. How about, if the Mars Society goes back Google-wise in time and dredges up any pre-internet proposals and schemes which have been allowed to sink out of sight. I've about "had it" with finding out my inspirational ideas are merely re-inventions of the wheel!

dicktice · 2003-11-19 18:34:48

Josh: Re. Any mutations would be science studies.
Aren't you being a little cavalier regarding self replication "mutations," in that you don't exclude "thinking machines," perhaps leading to self-consciousness, leading to boredom, down-tools, revolt and eventally, (gulp) takeover. . . ?

Josh Cryer · 2003-11-19 19:38:55

dicktice, it does make you wonder, doesn't it? I mean, though I don't consider the Apollo mission the main cause or builder of chip technology, I do totally realize that it was a contributor. Thinking about if we had such a base on the moon, it's a wonder where we'd be right now. For one, the base would've had potentially unsurpassed resources; R&D now costs millions, for a moon base, it would cost 'pennies'; such a base would've seriously pushed us into space drastically.

Also, I don't think auxon-like technology could ever reach intelligence; mutations would be at a small level; bots breaking but their internal programming adjusts so that they can still function marginally, etc. If intelligence occured (say several thousand machines configured themselves to create a working hive mind- a statistical improbablity to be sure), I think we'd need to be 'nice,' but surely it'd be something rather amazing to behold.

Josh Cryer · 2003-11-19 19:46:26

BTW, I've about "had it" too with finding my ideas have already been thought of or otherwise are thought up by someone else later on. The rail gun idea I thought myself was rather novel, and here I am reading a paper talking about using rail guns to distribute autons/auxons/self-replicating technology on the moon (and to other places in the solar orbit).

We humans are an interesting lot.

Palomar · 2003-11-20 06:30:31

BTW, I've about "had it" too with finding my ideas have already been thought of or otherwise are thought up by someone else later on. The rail gun idea I thought myself was rather novel, and here I am reading a paper talking about using rail guns to distribute autons/auxons/self-replicating technology on the moon (and to other places in the solar orbit).
We humans are an interesting lot.

*Hi guys: This is a really interesting thread in all respects. I like to consider the future of robots, and that "hive mind" potential is an intriguing concept. It all is. I think it's entirely possible they could "take over" at some point. We'd be wise to start programming in loyalty to the creator (humans) or something along that line...

There's a robot in Japan (?) -- I'm pretty sure -- which I saw a TV program about around 2000 or 2001. Scientists are speculating that this little 'bot has already achieved quite a level of sentience; it will look around the room, as if mildly studying its surroundings and appears to be thinking by itself when it's in "suspend" mode. I haven't heard anything more about it.

I "hear you" on the "that's already been thought of/realized?" matter. It happens in my philosophical studies all the time. Now I'm reading some of Marcus Aurelius' material...he lived 2000 years ago and I'm like, "Wow...I can relate!" I guess Solomon (another oldie...even older than Aurelius) was onto something when he said "there's nothing new under the sun" (as regards human nature anyway).

--Cindy

Free Spirit · 2003-11-22 20:12:40

*Hi guys: This is a really interesting thread in all respects. I like to consider the future of robots, and that "hive mind" potential is an intriguing concept. It all is. I think it's entirely possible they could "take over" at some point. We'd be wise to start programming in loyalty to the creator (humans) or something along that line...

Believe it or not, there are actually computer scientists out there who support building AI that will put the human race out of business. Marvin Minsky (one of the leading computer scientists) even went so far as to say that our AI creations will eventually "retire" the human race and that our extinction at their hands will be no big deal in the scheme of things. And yet people like Shaun accuse me of being some anti-human monster for opposing the reckless development of technology, as he so slightly hinted at in another thread.

Josh Cryer · 2003-11-22 20:53:50

It's a weird philosophical argument, really. I don't really care for strong AI, as they call it, because strong AI has the potential to enhance itself (humans can enhance themselves, but to only limited extents, and even then, it doesn't really mean enhancing their being, just how they interact with the world; ie, search engines make us more knowledgable, but that knowledge isn't always accessable).

So all my arguments here have been basically talking of technologies that we pretty much have complete control over. This way, if something goes wrong, there will be a human to blame, and not a machine.

But, if we were to talk about strong AI, I don't see it as much as a leap to say that robots would have a better survivablity chance; so if they did exist, they would be the ones "doing stuff" in space. And in that sense, since they'd be able to colonize... anywhere... we would be the minority. It's a scary thought, to be sure.

I feel though, that as long as human DNA exists (and all the knowledge along with it), the human race will never 'die.' Something to bork over.

Palomar · 2003-11-22 20:58:28

*Hi guys: This is a really interesting thread in all respects. I like to consider the future of robots, and that "hive mind" potential is an intriguing concept. It all is. I think it's entirely possible they could "take over" at some point. We'd be wise to start programming in loyalty to the creator (humans) or something along that line...
Believe it or not, there are actually computer scientists out there who support building AI that will put the human race out of business. Marvin Minsky (one of the leading computer scientists) even went so far as to say that our AI creations will eventually "retire" the human race and that our extinction at their hands will be no big deal in the scheme of things.

*Whoa.

I've not heard of Marvin Minsky before. Free Spirit, you've piqued my curiosity. I'll do a Google search for him in the next few days, but I'd like especially to read about this in detail. Can you refer me to a specific web link or magazine article (online or otherwise), in the event I miss something on Google?

Feel free to send me a Private Message with further information, if you like.

--Cindy

Josh Cryer · 2003-11-22 21:11:57

Another person you may wish to dig up is Ray Kurzweil. He's one of the pioneer "Singulartians" who believe that technology will advance so much in the next 30 years or so that computers will pretty much take over and we'll all live in virtual reality. Here is a link to his site, heh.

You will find some of Marvin's writings on that site. Ah, what the heck, here's a direct link for the lady.

All very odd (yet equally interesting) stuff that may be a wee bit off the topic of this thread.

Palomar · 2003-11-22 21:52:50

Another person you may wish to dig up is Ray Kurzweil. He's one of the pioneer "Singulartians" who believe that technology will advance so much in the next 30 years or so that computers will pretty much take over and we'll all live in virtual reality. Here is a link to his site, heh.
You will find some of Marvin's writings on that site. Ah, what the heck, here's a direct link for the lady.
All very odd (yet equally interesting) stuff that may be a wee bit off the topic of this thread.

*Aw, Josh, how nice of you! I'll look over the information. I'm totally unfamiliar with this stuff.

Thanks!

--Cindy

New Mars Forums

Announcement

#1 2003-06-03 08:14:11

Re: Robot Colony - Robots prepping for a human mission.

#2 2003-06-03 09:29:20

Re: Robot Colony - Robots prepping for a human mission.

#3 2003-08-08 12:13:42

Re: Robot Colony - Robots prepping for a human mission.

#4 2003-08-12 11:27:15

Re: Robot Colony - Robots prepping for a human mission.

#5 2003-08-12 23:04:46

Re: Robot Colony - Robots prepping for a human mission.

#6 2003-08-12 23:50:05

Re: Robot Colony - Robots prepping for a human mission.

#7 2003-08-15 15:37:59

Re: Robot Colony - Robots prepping for a human mission.

#8 2003-08-15 18:52:10

Re: Robot Colony - Robots prepping for a human mission.

#9 2003-08-15 23:02:57

Re: Robot Colony - Robots prepping for a human mission.

#10 2003-08-15 23:42:44

Re: Robot Colony - Robots prepping for a human mission.

#11 2003-08-16 01:48:59

Re: Robot Colony - Robots prepping for a human mission.

#12 2003-08-16 03:01:13

Re: Robot Colony - Robots prepping for a human mission.

#13 2003-08-19 04:00:48

Re: Robot Colony - Robots prepping for a human mission.

#14 2003-09-08 23:21:19

Re: Robot Colony - Robots prepping for a human mission.

#15 2003-11-19 05:37:18

Re: Robot Colony - Robots prepping for a human mission.

#16 2003-11-19 18:22:16

Re: Robot Colony - Robots prepping for a human mission.

#17 2003-11-19 18:34:48

Re: Robot Colony - Robots prepping for a human mission.

#18 2003-11-19 19:38:55

Re: Robot Colony - Robots prepping for a human mission.

#19 2003-11-19 19:46:26

Re: Robot Colony - Robots prepping for a human mission.

#20 2003-11-20 06:30:31

Re: Robot Colony - Robots prepping for a human mission.

#21 2003-11-22 20:12:40

Re: Robot Colony - Robots prepping for a human mission.

#22 2003-11-22 20:53:50

Re: Robot Colony - Robots prepping for a human mission.

#23 2003-11-22 20:58:28

Re: Robot Colony - Robots prepping for a human mission.

#24 2003-11-22 21:11:57

Re: Robot Colony - Robots prepping for a human mission.

#25 2003-11-22 21:52:50

Re: Robot Colony - Robots prepping for a human mission.

Board footer