In CRN's construct of a desktop nanofactory, products are built by putting together combinations of predesigned nanoblocks. This is intended to maximize the latent innovation potential in the widespread distribution of low-cost (or free), albeit technically restricted nanofactories. It also could prevent illicit, unwise, or malicious product development.
Product design will be made simple by CAD (computer aided design) programs, so simple that a child can do it—and that’s no exaggeration. New product prototypes can be created, tested, and refined in a matter of hours instead of months. No special expertise is needed. Just imagination, curiosity, and the desire to create.
Functionality must be contained within small blocks, which must then be fastened together by the trillions to make a finished product. In our initial design, the blocks are only 200 nanometers on a side. However, this is big enough to hold a small CPU or a motor. Larger functionality can be split between blocks—power, force, and signal can all be transferred between the blocks by simple, efficient interfaces. A product that needs a supercomputer or a powerful motor can combine smaller computers or motors. This splitting will add some inefficiency, but not enough to seriously impair product function.
When combined with automated verification of design safety and protection of intellectual property, this approach will open up huge new areas for originality and improvement while maintaining safety and commercial viability.
Working with nanoblocks, designers of all ages, nations, and backgrounds can create to their hearts’ content. The combination of user-friendly CAD and rapid prototyping will result in a spectacular synergy, enabling unprecedented levels of innovation and development. Among the many remarkable benefits accruing to humanity from nanofactory proliferation will be this unleashing of millions of eager new minds, allowed for the first time to freely explore and express their brilliant creative energy.
I'm seeing nano-scale Capsela, Fischertechnik, Electronic blocks, logiblocs, or maybe even Legos.
Is this the notion of which you are speaking?
Posted by: Bravo Romeo Delta | March 21, 2004 at 01:35 PM
Yes, that's it. Except, at least in the beginning, all the nanoblocks will be precisely the same dimensions, perhaps 200 nm cubes. That will make the initial CAD programs easier to write. It's also likely that there will be several design levels--something like Beginner, Intermediate, and Advanced--with greater varieties of blocks available at each level, for making more sophisticated designs.
Posted by: Mike Treder, CRN | March 21, 2004 at 02:05 PM
So, essentially, closest to Capsela in conception. So, the actual business of doing 'complex' things, like creating new sorts of nanoblocks could be shopped out to actual designers, while the remainder of the design could be done at home, or based on whatever whim one had in mind.
Good thought.
Posted by: Bravo Romeo Delta | March 21, 2004 at 02:43 PM
Speaking as a mechanical designer, building macro-scale mechanical devices out of identically sized and proportioned nano-scale blocks could be a bit challenging. Linear motion can be accomplished easily in such a system, but continuous rotory motion? You'll need some blocks of variable size and shape; Brackets, for instance, with a lockable Stewart platform between two fastening surfaces, would be one potential solution. Ideally your nano-factory, in addition to the standard functional blocks, should be able to generate inert structural blocks with arbitrary geometry.
Large objects capable of continuous rotary motion require large rotary bearings, at the least large, very smooth surfaces of radial symmetry. You're not going to build such surfaces out of rectangular blocks.
An alternative approach, of course, is some variation on the utility fog idea. What utility fog loses in sheer performance, it certainly makes up in versitility, and a bit of the fog would go a long way towards easing the design of less flexible machines. A programable filler material.
Then there's the need for sealable surfaces, often flexible. The enviroment is going to be full of things you just do not want getting into the works. This is just surface, not volume, so it might be feasible to do post-assembly, utilizing some kind of chemical depostion nano-bots, which could deposit a variety of sealing and colorant materials over the desired surfaces, not necessarilly controlled right down to the nano-scale.
Designing a reasonably complete set of blocks is no trivial task. It would be eased greatly by having some blocks which were more parametric in nature, rather than fixed.
Posted by: Brett Bellmore | March 21, 2004 at 03:03 PM
Brett, I did think about large rotating interfaces/bearings. See for example section 3.2.2 a few paragraphs in. http://www.jetpress.org/volume13/Nanofactory.htm#s3.2.2
The same section proposes diamond or graphite curtain walls as a way of sealing large pipes.
And I did specify (somewhere) that the blocks could be truncated, as long as the handling robotics could still grab them.
Keep in mind that I wrote the paper mainly to show what could be done with a very primitive version of molecular manufacturing that we could design today (aside from the mechanochemistry, which may take only a few million US$ to figure out) and that I could describe in an 80-page paper. A real nanofactory, even an early version, will probably have somewhat more flexibility.
Posted by: Chris Phoenix, CRN | March 21, 2004 at 10:39 PM
What molecular CAD program go you recommend? I have used RASMOL and am looking at PYMOL, MMTK, and Fungimol. I am primarily interested in dry diamondoid nanotech. I don't need a lot of bells and whistles. I don't need special protein or DNA functions. I would like something that is easy to edit, point and click adding and removal of atoms, moving atoms. I want to design machinery using mostly nanotube struts, buckyballs, diamondoid casings. Working with simple carbon structures I wouldn't need a lot of hand holding for the bond angles or strengths, MM2 or MM3 would be fine. But if it had nothing at all, that would be okay too. I think I know enough chemistry to do it manually. And of course, I don't have a lot of money to spend.
Posted by: Mike Deering | March 22, 2004 at 07:54 AM
Mike, I don't know what to recommend; I haven't looked at molecular CAD for a few years. When you find out, please let us know!
Chris
Posted by: Chris Phoenix, CRN | March 22, 2004 at 01:09 PM
okay.
Posted by: Mike Deering | March 22, 2004 at 10:29 PM
My one concern about re-usable nanoblocks, which aren't covalantly bonded, is that you might very well be able to design a device out of them, which didn't use mechanochemistry, to disassemble artifacts built up out of them, and make copies of itself. (Somebody already suggested this could be a problem.) Not quite as serious a threat as grey goo which could disassmble YOU for raw material, but still rather annoying.
I would suggest that any block which could contain sufficient processing power should be "keyed" to the nanofactory which made it, such that it can't be re-used without an authorization code supplied by that nanofactory. This would allow you to re-use your own nanoblocks, without their being fodder for any self-replicating device that wandered by. Such a thing would still be able to assimilate the stupider blocks, but would be limited by not being able to assimilate some essential classes of them.
Posted by: Brett Bellmore | March 23, 2004 at 08:03 AM
Brett,
That was my idea and I would like to purpose the term "gray gobblers" for self-replicators that reuse pre-existing nano-blocks in the environment as their source of raw materials.
I personally do not think that they will be a major problem unless they start housing advanced AI's.
Posted by: jim moore | March 23, 2004 at 05:57 PM
An architect's version of the vision:
Poway Man Designs Qubits
http://www.nctimes.com/articles/2005/09/06/front
or
http://www.nctimes.com/articles/2005/09/06/news/inland/9505193122.txt
Fresh Patents
http://www.freshpatents.com/Mark-Allan-Burginger-Poway-invdirb.php
Art
http://www.Burginger.com/MODULAR/GEARS_2.mov
Posted by: Mark Burginger | October 09, 2005 at 11:31 AM