• Google
    This Blog Web

October 2011

Sun Mon Tue Wed Thu Fri Sat
2 3 4 5 6 7 8
9 10 11 12 13 14 15
16 17 18 19 20 21 22
23 24 25 26 27 28 29
30 31          

RSS Feed

Bookmark and Share

Email Feed

  • Powered by FeedBlitz

« Barriers are Coming Down | Main | Russian Kids Doing MNT? »

March 28, 2004


Feed You can follow this conversation by subscribing to the comment feed for this post.


Yes, the upheaval will be truly impressive.

The guy who makes a "fab in the box" is going to make a hell of alot of money. This is definitely doable and actually could be done using "conventional" self-assembly chemistry. The first step in this direction is "roll-to-roll' or web processing for making flat-panel displays. next will be the addition of self-assembly chemistry.

I have friends who are chemists (who do NOT believe in Eric Drexler's nanotech, by the way) who say that this can be done in about 10 years or so.


looks to me the transition to the nano-era is going to go much more smoothly than previously thought; sure, america looks to loose in the ten years, but at least the whole world does'nt self-destruct in one year some time soon

Rocky Rawstern

And it looks to me that here again is yet another route to "nanotech." I doubt that this will be the last. Which only lends credibility to the notion that we'd better start paying attention, now.

Bravo Romeo Delta

It may very well mimic the "upheaval" that went along with the internet. From nothing to ubiquitous without the date of change getting put into anyone's daytimer.

Chris Phoenix, CRN

Kurt, please say more about how self-assembly can lead to "fab in a box." How do you program in the complexity? With directed assembly/chemistry (either wet or dry), you can download the complexity in the instruction stream, rather than building it physically into the manufacturing infrastructure.

Of course, relatively simple things can be quite valuable: displays, RAM, medicines. But if you can't build at least a GB of complexity into your product, you'll be missing a whole universe of products.

But to put gigabytes or even terabytes of complexity into the starting chemicals, and then let them self-assemble into the desired configurations and features... I don't understand how this is feasible (especially if the product has features of orders-of-magnitude different sizes), but I'm ready to learn.


Chris Phoenix, CRN

...Well, I'm surprised. I was going to write that David was wrong about the gentler transition, because 3D printing couldn't make manufacturing cheap, because a manufacturing system couldn't duplicate itself in a reasonable time. So I went and looked... and found that color printer inkjet technology might deposit ~3 cc per minute, enough to build ~10 kg per hour, which certainly implies building its own mass in a remarkably short time.

Whether or not it's nanotech, if someone works out a way to make a printer print a printer in a day (which also implies motors and semiconductors and kinematics), then we could see a manufacturing revolution. I don't immediately know how to build long-lasting bearings, and several other technical details, but I don't see any obvious showstoppers.

And the engineering would be immediately accessible to people with macro-scale intuitions. (Nanoscale engineering will be more accessible than many people think, but that's another story.) I'm picturing a contest developing between college-level mechanical engineering and product design departments: who can build a ballpoint pen? A cordless phone? A general snap-fit system for assembling larger products with minimal labor (or no labor, if it's roboticized)? A general fractal truss for maximum compressive strength? And so on.

Maybe I'm running ahead of things here. But this looks potentially pretty big. I hadn't followed up on dot-matrix rapid prototyping before this because I thought the materials were weak, the parts needed post-processing, and most of all, the systems were too slow to usefully duplicate themselves. But if I was wrong about those limitations... perhaps I should look into this a bit more. At least it will be a useful comparison with molecular manufacturing.


Brett Bellmore

I don't think inkjet could make a complete manufacturing system, but it could probably get you darned close. A partially self-reproducing factory, which couldn't quite make all it's own components, would still be very useful, and suplemented with some other technology could make a self-reproducing whole.


I think the upheaval will be more significant than compared to the upheaval that the internet caused.. It will be more like the first industrial revolution, wich changed the economy beyond recognition in just a couple of years. Where we had essentially a home industry building everything by hand and a very little output of luxury stuff, all of a sudden there were factories, bringing mass-produced -and so magnitudes cheaper- products for the first time in the financial reach of the 'common man'. But those common men and woman lost their home-industry jobs and had to go work in said factories...
How will economy change this time? Massive lay-offs in a very short period... Essentially everbody working on the shop-floor will become redundant virtually overnight. Maybe back to human-service industry jobs? Will designers/programmers be the only people working in 'industry' after this transition, and the 'unskilled' looking after their kids? Hmmm... But most of the work can be done at home (via digital networks)

Are there any studies done on a future with 'labourless' manufacturing?

Richard Jones

I'm not convinced that the dislocation caused by a move to "laborless manufacturing" will be that great, simply because this shift is well under way already. Here in the UK, only 15% of jobs are in manufacturing; this fraction has been falling steadily for at least 20 years. The UK may be slightly further down this path than other advanced economies, but this is a universal trend.

To see why the fact that material things already cost virtually nothing makes surprisingly little difference to the way our economies work, ask yourself the next time you buy, say, a $100 pair of running shoes, how much of this value actually goes to the factory that makes them in Vietnam or wherever. $20 maybe? The great majority of the value added to our goods already derives from services - design, marketing, retailing and distribution and the like, and this will remain true even when the actual cost of manufacturing really does become negligible.

There is an interesting question about what difference it makes if the capital cost of setting up a manufacturing plant falls - if you really can replace the $2 billion it costs to set up a fab line for LCD displays, say, with a few ink-jet printers. Does this mean everybody is going to make everything they need at home? I'm not sure. Tom Coates wrote an article called "The mass amateurisation of nearly everything" based on the observation that something very similar has happened in the media recently. I could set myself up to make professional quality TV programs for less than $10,000 now, and as we see, everyone can write a blog and reach a worldwide audience for virtually nothing. Does this mean that printed media have died off and the film and TV industries have withered away? No, and I don't think they will, because consumers are odd and illogical creatures who remain more influenced by irrational things like fashion and brand values than by anything substantial like design quality or functionality.

If decentralised manufacturing really does become a reality, I think it's much more likely that it will take place not in the home, but in a store in the mall, where what is an intrinsically almost worthless product will be imbued with value by association with that store's brand values.

Karl Gallagher

Imagine the shoe salesman of the future. He comes to your house, spends an hour interviewing you about your likes and dislikes, watches you walk around the house and through the park, analyzes the terrain, talks to some of the people in your social circle to learn the fashion expectations, then writes up a detailed specification for a set of shoes to meet your every need. That gets translated into a design by another (lower-paid) professional, and then your home fab spits out six pairs of shoes that cover your every need, along with a data disk so you make make replacements when travelling.

3 days of shoe expert: $500
1 day of design expert: $100
6 pairs of shoes: $0.50

Now imagine replicating that across every object you touch in a day.


Nano-imprint lithography is a first step towards a "fab in a box". Those of you in the industry know that photo-lithography is the most expensive process and equipment in the fab. The current 157nm DUV system range in price from USD10-20million. The nano-imprint systems are several hundred thousand and are the size of a desk. They have pattern feature sizes down to 10-20nanometers. Of course, nano-imprint cannot match photo-litho on through-put, but the time will come.

Nano-litho systems are going into the MEMS markets where the economics are totally different than those of semiconductors (the process equipment must be cheap). Of course, being able to grow the devices using self-assembly chemistry rather than litho-patterning is the desired approach. There is lots of work going on in organic semiconductor and display materials. The reason is that these materials can be grown using self-assembly chemistry rather that the depositon, pattern,and etch process cycles that are currently used to make semiconductors.

Chris Phoenix, CRN

Richard: interesting point about the economics. But it's not just laborless manufacturing--it's also distributed (no shipping) and fast (no storage).

Rxke: Marshall Brain has written about robots taking over many service jobs as well as manufacturing. http://marshallbrain.com/robotic-nation.htm


Tom Mazanec

How much to have a scanner scan my feet and program an "almost perfect" pair of shoes which would be still far better than anything I can buy today?

The comments to this entry are closed.