Engineering Microswarms | Eclipse Phase

Sun, 2011-12-11 22:20

#2

OneTrikPony OneTrikPony's picture

Re: Engineering Microswarms

My understanding is that the intent was for engineer swarms to use the time's listed for fabbers on page 285. I've always thought that was unrealistically fast (but I don't really understand the principles of self assembly), so in games I play in I lobby for a house rule that also accounts for the size category on page 297. Typically we assign a value of 0 to nano sized gear and increase the time by 1 for each successive size increase; Nano 0, Micro 1, Mini 2, ect. It's not quite balanced and still miraculously speedy in most cases. You can fabricate half a cubic meter of aerogell in 5 hours (probably too long) and a synth morph of about the same volume in 25 hours. But its better than being able to turn out a Liquid steel morph every 5 hours. [edit] OOPS sorry, my point was that you might want to talk to your GM about using the size catagory table to judge the time it takes your engineers to build something since the really only build basic non complicated structures wich would always have a cost of trivial unless materials are scarce for some reason.

—

Mea Culpa: My mode of speech can make others feel uninvited to argue or participate. This is the EXACT opposite of what I intend when I post.

Sun, 2011-12-11 22:33

#3

Norade

Re: Engineering Microswarms

We pretty much went with a swarm working as quickly as a reasonably competent five man work crew. I think that seems a bit fast, but like you said it's hard to picture exactly how swiftly they should get things done. I'm still hoping that one of the members here who seems to know a ton about nanoswarms steps in and says something sciency. xF

Sun, 2011-12-11 23:45

#4

Axel the Chimeric Axel the Chimeric's picture

Re: Engineering Microswarms

That seems incredibly fast for a nanoswarm, but not so much for a microswarm. I would also imagine that even nanoswarm constructors are probably not able to build extremely complicated objects outside of a nanofabber. This makes them great for "growing" buildings, or pretty much any structure that doesn't require nano-scale precision, like low-tech technology. Again, this is great for growing buildings, or for rapidly building simple tools in a survival situation. It's also good for infiltrators, as long as you have a fairly simple tool set that you need (a traditional firearm definitely qualifies under this definition). Not so great if you're using high tech. I'm not so sure in terms of speed, myself. I've got things that "seem" right but I'm really not qualified. I leave such things to someone more qualified than myself. I'd imagine a 15x15x10 single room shelter composed of sturdy plastics could be assembled in less than a day, though.

—

[img]http://img813.imageshack.us/img813/982/exhumanbar.jpg[/img] [img]http://img804.imageshack.us/img804/4473/scumbar.jpg[/img] [img]http://img818.imageshack.us/img818/1396/gatecrasherbar.jpg[/img] [code][@-rep +1, f-rep +2][/code]

Mon, 2011-12-12 01:08

#5

OneTrikPony OneTrikPony's picture

Re: Engineering Microswarms

This is one of the things I wish would be fleshed out more in future supplements. (except that I suspect future supplements would continue to disagree with me concerning the time frame.) The problem is that what Luddites like me tend to think of as "nanotech" is really a massive, cross-discipline field of research. It's way to complex to be covered by; "one hour per cost category" with out feeling like the equivalent of magical conjuration. If you would like to dig into it a bit I recommend reading Robert Freitas and the institute for molecular manufacturing; Links below. http://www.rfreitas.com/ http://www.imm.org/ I only understand about three words out of ever nine but what I've read so far causes me to have serious issues with things like "engineer" and "protean" nano assembers that apparently work anywhere. For instance; how do you get your nanites to manipulate carbon or iron molecules in the presence of oxygen without building bulk rust and slag[b]?[/b] Or how do these swarms produce the rarified environment that is usually needed to manipulate single molecules or volitile and reactive substances[b]??!![/b] How does anything get built at the nearly 0 kelvin temperatures out in space[b]?[/b] I'm not arguing against the gear or bitching about "realism" but some more technical information and jargon would make me feel less like it's just RP magic. If the Devs are going to spend rez points on any part of the game mechanics I wish they would start with the Nanotech, cauz those rules suffer seriously from lack of resolution.

—

Mea Culpa: My mode of speech can make others feel uninvited to argue or participate. This is the EXACT opposite of what I intend when I post.

Mon, 2011-12-12 07:43

#6

Xagroth

Re: Engineering Microswarms

CMs seem to use "bush fingers" of sorts, while the nanoswarms use semi-autonomous nanoscale robots of some kind. It is not really hard to assume that both are able to generate a "bag" that allows them to work in vacuum, sealed and protected conditions if they need to build stuff bigger than themselves. However, while a Cornucopia Machine has a limit regarding its top speed when assembling, a swarm can work faster just with the owner increasing its size, then recycling the robots. Considering how much more powerfull swarms are, it is not hard to see that the increase in abuse potential is not one step, but several above the CMs, meaning that even in places like Extropia they might be heavily controlled. That means that the investment/output formula for Player Characters tends to go into the Cornucopia Machines rather than the nanoswarms for goods production, while the nanoswarms and microswarms are much more useful as a weaponized tool (able to disable any morph, exoskeleton or weapon in a few rounds unless the target is equipped with its own protection system against nanotech attacks). Of course, the main disadvantage of weaponized swarms comes from them being purposely unshielded against EMP grenades (as part of the "keep this shit under control" policy that differenciates the TITAN swarms from the Transhuman ones). If you want rough and dirty estimations of building times for nanoswarms, consider that only when you talk about assembling items bigger than a car the nanoswarms give a time advantage. As for the consideration of complexity, materials, and size interrelationated I posted somewhere in the forum a rough estimation of those factors.

Mon, 2011-12-12 10:22

#7

Axel the Chimeric Axel the Chimeric's picture

Re: Engineering Microswarms

Cornucopia machines work so well, down to the nanoscale, because they create a sealed environment within which the nanobots and nanoscale manipulators can work. They're an incredibly effective device as a result. Engineer swarms working outside said environment have to be considerably less... Choosey. That's why engineering swarms are microswarms, not nanoswarms. They construct simple structures, likely taken from a specified feedstock or carved out of something, rather than having the given type of material assembled by them. While it's not impossible for engineer swarms to build using materials from the local environment, that probably makes the process take significantly longer, especially if the structure must remain sound and stable. They bypass a lot of nano-related concerns by simply not quite having to work on that scale. One exception might be in laying nanotube wire, but that's another story. I can see microbots merging together and forming a mobile factory on a tiny scale, into which the others bring feedstock, which it converts into the necessary form (carbon into nanotubes, spitting aerogel, etc.) while moving along to lay it. They can also probably manage the rather amusing spectacle of moving pre-assembled components of a structure into place and laying polymers to bind it there. Protean swarms, meanwhile, probably do assemble what you might call a Cornucopia-In-A-Bag, as is suggested. They assemble an extremely thin sheath using feedstock included with them that surrounds the area they're applied to. They then proceed to rapidly transform it into their work environment and set about consuming everything inside to generate the required item. When it's done, the owner breaks the extremely thin sheath, pulls out the item, and they're good. This makes protean swarms, frankly, largely useless for mainstream items inside a city. Sure, they can grow you a car, but it'll take ages as they set up the sheath and prepare the materials - it's easier just to rent the time from a large-scale CM machine. When you're stuck in the desert and you need a rapidly prepared canister of fuel or a vehicle to get you back to town without walking, you'll be thankful, even if you have to wait a good number of hours. They're also excellent for infiltrators, who can simply bring an unassuming cluster of nanites kept in a shielded pocket inside their concealed cybernetic right arm. They toss it on the ground and it turns what was formerly a metal sculpture and part of the floor into a gun, some ammunition, and a few other useful tools.

—

[img]http://img813.imageshack.us/img813/982/exhumanbar.jpg[/img] [img]http://img804.imageshack.us/img804/4473/scumbar.jpg[/img] [img]http://img818.imageshack.us/img818/1396/gatecrasherbar.jpg[/img] [code][@-rep +1, f-rep +2][/code]

Mon, 2011-12-12 11:44

#8

OneTrikPony OneTrikPony's picture

Re: Engineering Microswarms

Axel the Chimeric wrote:
... That's why engineering swarms are microswarms, not nanoswarms. They construct simple structures, likely taken from a specified feedstock or carved out of something, rather than having the given type of material assembled by them. While it's not impossible for engineer swarms to build using materials from the local environment, that probably makes the process take significantly longer... Protean swarms, meanwhile, probably do assemble what you might call a Cornucopia-In-A-Bag... This makes protean swarms, frankly, largely useless for mainstream items inside a city. Sure, they can grow you a car, but it'll take ages as they set up the sheath and prepare the materials...

This is sort of similar to my current understanding of how the tech is intended to work eventually. I'd like to have this represented in the time it takes but (unfortunately) protean swarms are noted to use the same time frame as Fabbers. (page 329) the difference of my view is that proteans needn't build a macro structure "CM machine in a bag" but would need to constantly fabricate and recycle shielding for the specific "micro work sites". Also, my understanding of prospective technologies in "Molecular [i]Self[/i] Assembly" is that the bots are intended to become part of the manufactured item themselves similar to the concept of [url=http://en.wikipedia.org/wiki/Claytronics]claytronics or Catoms[/url] Obviously EP nano needs additional feedstock but that's the reason that Swarms are One Shot items. Still I'm dumbfounded by the apparent level of complexity represented by a swarm. It's equivalent to Prokaryotic life forms. I imagine it would require Dozens of different types of individual bots; Dissasemblers, "lego stackers" transporters, recyclers, re-fuelers maybe...

Xagroth wrote:
Of course, the main disadvantage of weaponized swarms comes from them being purposely unshielded against EMP grenades

RAW I can't find any info on EMP effecting swarms. Do you have a reference for that? Technically tho, I thought that nanotechnologies were impossible to shield from EMP or even just powerful magnetism due to their size.

—

Mea Culpa: My mode of speech can make others feel uninvited to argue or participate. This is the EXACT opposite of what I intend when I post.

Wed, 2011-12-14 08:09

#9

Xagroth

Re: Engineering Microswarms

I think CM works with "bush fingers", allowing them to "print" the item in the nanoscale level, while the swarms need to move around (which includes going to the "feed block" to gather some materials all the time), which slows down the process. The main advantage of swarms Vs CM is that they are not fixed to a predeterminated number of bots ("fingers" for the CM), so if you want to make a really, really, really big item (like a spaceship), once you have the raw materials you can make tons of bots to speed up the process. EMP Grenades: pg 340 of the corebook (3rd printing) Swarms: Pg 329 (corebook 3rd printing):

Quote:
Area-effect weapons, plasma rifles, and fire inflict 1d10 damage, plasma grenades do full damage. [b]EMP weapons (p. 341) are very effective against swarms, inflicting 2d10 + 5 damage and a –10 modifier to all tests due to their damaging effects on the swarm’s communication abilities until repaired[/b].

Wed, 2011-12-14 11:54

#10

Axel the Chimeric Axel the Chimeric's picture

Re: Engineering Microswarms

Swarms probably form "work chains", passing materials along a highway created by a long series of small bots. The workforce of bots is probably divided amongst the tasks of delivering materials, forming structures, and processing materials to be delivered.

—

[img]http://img813.imageshack.us/img813/982/exhumanbar.jpg[/img] [img]http://img804.imageshack.us/img804/4473/scumbar.jpg[/img] [img]http://img818.imageshack.us/img818/1396/gatecrasherbar.jpg[/img] [code][@-rep +1, f-rep +2][/code]

Wed, 2011-12-14 14:30

#11

OneTrikPony OneTrikPony's picture

Re: Engineering Microswarms

Xargoth; Thanks for the Quote! :D I really need to buy a new book I'm still using my first edition print copy sometimes. Guess I'll tape in another post it note.

—

Mea Culpa: My mode of speech can make others feel uninvited to argue or participate. This is the EXACT opposite of what I intend when I post.

Thu, 2011-12-15 08:26

#12

Xagroth

Re: Engineering Microswarms

OneTrikPony wrote:
Xargoth; Thanks for the Quote! :D I really need to buy a new book I'm still using my first edition print copy sometimes. Guess I'll tape in another post it note.

You can get a digital copy here (uploaded by Posthuman Studios) in the meantime: http://www.demonoid.me/files/details/2625856/004626639420/

Thu, 2011-12-15 16:45

#13

OneTrikPony OneTrikPony's picture

Re: Engineering Microswarms

Thanks. I bought the third printing PDF I just don't use it very often cause I usually have sunward on the right screen and what ever I'm working on on the left screen and Dead Trees open on my desk. I need a new dead tree copy.

—

Mea Culpa: My mode of speech can make others feel uninvited to argue or participate. This is the EXACT opposite of what I intend when I post.

Sun, 2011-12-25 06:01

#14

Arenamontanus Arenamontanus's picture

Re: Engineering Microswarms

(sorry for the late intervention) I discussed the whole CM vs swarm issue with Eric Drexler a few weeks back, and he argued that swarms are really lousy for manufacturing. When you make things you want to put parts together, so you don't want to move them far and you want to maintain good control over where everything is. A CM would not just have a controlled internal environment, it would also use convergent assembly - nanodevices put together tiny pieces that are picked up by microdevices that put them together into mini-pieces that are then put together by macrodevevices into the finished product. The nice part is that the systems can all world in parallel. Fractal robot arms are great for this. This convergence is hard to do with a swarm - no great positional control, raw materials and other things have to be moved through the environment, no big structures for handling larger parts and so on. Also, they need energy to break or make chemical bonds: it is tough for a swarm to carry that much, it needs to get it from the environment (tapping into power cables or burning materials?) - and conversely, cooling is slapdash. The only thing swarms are good for is control or monitoring of entire volumes - guardians, saboteurs, gardeners, tracking and so on. So I would argue that proteans and engineering swarms are one level slower than the CM, at least. They are convenient for transporting and sneaking in, but weak for manufacturing.

—

Tue, 2011-12-27 17:37

#15

OneTrikPony OneTrikPony's picture

Re: Engineering Microswarms

Arenamontanus wrote:
(sorry for the late intervention) I discussed the whole CM vs swarm issue with Eric Drexler a few weeks back,

[b]![/b] You talk to Eric Drexler? Like, just sit and chat? I'm so friken Jealous right now. (but also happy that you bother to hang out here. :) )

Quote:
Also, they need energy to break or make chemical bonds: it is tough for a swarm to carry that much

So, would it be accurate to say that the minimum energy requirement to make a given object is twice that contained in all of it's Covalent, Ionic, and Metallic bonds? (1/2 the energy goes to break down the source material and the other half to re-assemble.) Also, have you ever run across a proposal that waste heat from the assembly process itself might be channeled back into the system to power the assemblers? (I think it might have been Drexler who proposed that but now I can't find the reference in my notes.)

—

Mea Culpa: My mode of speech can make others feel uninvited to argue or participate. This is the EXACT opposite of what I intend when I post.

Thu, 2011-12-29 07:36

#16

Arenamontanus Arenamontanus's picture

Re: Engineering Microswarms

OneTrikPony wrote:
Arenamontanus wrote:
(sorry for the late intervention) I discussed the whole CM vs swarm issue with Eric Drexler a few weeks back,
[b]![/b] You talk to Eric Drexler? Like, just sit and chat?

You know, it is a good thing we have offices on opposite ends of the building. Whenever we meet at the coffee machine we get stuck in fun discussions instead of work. (This is one of the biggest perks at being at a place like Oxford - a sizeable fraction of the Cool People of the world are around all the time, and you can talk to them.)

Quote:

Quote:
Also, they need energy to break or make chemical bonds: it is tough for a swarm to carry that much
So, would it be accurate to say that the minimum energy requirement to make a given object is twice that contained in all of it's Covalent, Ionic, and Metallic bonds? (1/2 the energy goes to break down the source material and the other half to re-assemble.)

That would be the brute force approach, very energy expensive. A lot of bonds are actually already of the right kind, so you just need to move them to new places. If you have some petroleum-like substance and make diamond for example, you will not need to break many carbon-carbon bonds. Instead you will break all hydrogen-carbon bonds and replace them with carbon-carbon bonds (plus a lot of hydrogen-hydrogen bonds). If you have a feedstock that already contains diamondoid substructures you need to break even fewer bonds. Eric said that you can really get the energy costs (and effort) down by having the right kind of intermediary feedstock. Imagine molecular lego pieces that can easily be snapped together or recycled, but take much more energy to make. So the habitat or city manufacturing complex make them in big volumes and pipe out feedstock, while consumer CMs mainly assemble and disassemble objects made up of these pieces.

Quote:
Also, have you ever run across a proposal that waste heat from the assembly process itself might be channeled back into the system to power the assemblers?

I don't think this works. Waste heat is not very good for powering anything (hence the waste in the name), due to the laws of thermodynamics. You give the assembler quality energy in the form of electricity, and it gets converted into heat. Turning some of the heat into electricity is doable, but you get a huge loss of efficiency. On the nanoscale the heat is the shaking that occurs when stuff snap together or apart. If you do things slowly and perfectly it is minimized.

—

Sat, 2012-01-07 20:03

#17

Lord High Munchkin Lord High Munchkin's picture

Re: Engineering Microswarms

One thing to remember is that large automated factories can churn out goods far faster and more cheaply/efficiently than nanotech. Anything repetitive and/or using many identical parts should always be simpler to make in a factory (although their products could also be used as feed-stock or components by smaller assembly systems and methods). Why make a spaceship with swarms when a series of automated factories in orbit could do it more easily? I miss Oxford, sometimes.

Mon, 2012-01-09 19:34

#18

King Shere King Shere's picture

Re: Engineering Microswarms

Lord High Munchkin wrote:
One thing to remember is that large automated factories can churn out goods far faster and more cheaply/efficiently than nanotech. Anything repetitive and/or using many identical parts should always be simpler to make in a factory (although their products could also be used as feed-stock or components by smaller assembly systems and methods). Why make a spaceship with swarms when a series of automated factories in orbit could do it more easily?

No. I think you are mistaken. Or you are not using the abbriviation "Nanotech" to for Nanotechnology.

Core rule book wrote:
"Nanotechnology is the precise manipulation of matter at the atomic level, typically using millions of microscale nanomachines. Nanotechnology transformed manufacturing, enabling new techniques and materials." page 326 "Basic nanotechnology is exceedingly widespread and used throughout the solar system, serving as the primary method for manufacturing for decades." page 326 .

Thus I expect the sol systems production lines & factories to use nanotech manufacturing techniques , 1000+ fabbers (factory) vs 1 fabber (field agent/individual designer). [b]How "small" a machine with the abbreviation nanobot is (depends on what it is)[/b]. A swarmanoid morph, its individual robots are insect sized (centimetre sized) page using nanocopter fan blade for uplift. page 144 Swarm nanobots are the size of microbes (microscale) naked to the human eye. page 328 Swarm micromachines are the size of small insects (centremeter scale).page 328 a Transhumanity "Nanobot" is thus= centimetre sized or micro sized machines , not nanosized. Despite eventual nanotech & nanobots tags. Our modern technology (not EP) have made Molecular machines ( nanosized) they are also known as nanobots. http://en.wikipedia.org/wiki/Molecular_machine http://en.wikipedia.org/wiki/Nanocar http://nanotechweb.org/cws/article/tech/38953 This article left me with the first impression that EMP might not work on nanoscale sized devices.

—

[img]http://boxall.no-ip.org/img/tc_userbar.jpg[/img] [img]http://boxall.no-ip.org/img/exh_userbar.jpg[/img] [img]http://boxall.no-ip.org/img/argo_userbar.jpg[/img] [i]"To find fault is easy; to do better may be difficult."[/i] Plutarch

Tue, 2012-01-10 00:36

#19

Lord High Munchkin Lord High Munchkin's picture

Re: Engineering Microswarms

King Shere wrote:
No. I think you are mistaken. Or you are not using the abbriviation "Nanotech" to for Nanotechnology.

You are correct that I was not using "nanotech" in its strict sense... but merely as "not macro". If you only need joined metal, a welding robot can do it too. Simple technology might be all that is required.

Tue, 2012-01-10 09:40

#20

Arenamontanus Arenamontanus's picture

Re: Engineering Microswarms

When you build a spaceship you likely use a series of assembly steps. Fabbers make different kinds of parts - often with great specialisation: one fabber makes lots of structural components, another lots of hull, a third sensor equipment and so on. This specialisation allows them to have pretty high throughput (they need just a certain kind of raw materials and can be optimized in making their product). Then robots or frames not unlike 3D printers pick up the parts and position them, where they are attached (not just by welding, but through matching surfaces that lock together down to molecular precision in many cases, super "glue" that integrates with the material or even nanoswarms finding gaps and filling them). In many cases this happens hierarchically - a cluster makes components that are fitted together into something larger, which is then linked with other parts made at the same time. Very parallel and requiring some tricky logistics software and engineering oversight. A general purpose fabber is useful but fairly slow. It can make anything but likely has a hard time really optimizing for your particular product. Fabbing a speciality fabber might sometimes be smarter. (Of course, this leads to the situation that many engineering sites are littered with speciality fabbers that might be hacked or repurposed for new uses... "Old man Zhang figured out how to get the dome plate makers to produce ultra-stressed diamond. Looks normal, but is a decent explosive when triggered. The funniest part is that he did it mainly as an exercise in his project to have them make alcohol...")

—