Manufacturing on Demand: The Future Is Now

Computer and internet technology is bringing big changes to daily life: music distribution without the need for big physical production plants, electronic book publishing, and even what the continuing advances in computer technology will mean to computers themselves. All of these changes really come down to one central idea: increasingly the real product is information, not a physical object built around that information. Computers and the Internet together make it so much less expensive to transmit that information that the whole business models of many industries, like record and book publishing, are dissolving as the physical model is replaced with an information model.

Still, sometimes we need physical objects: you can't sit on an information chair, and you can't eat with the binary description of a fork. But imagine if there were some way to automatically turn that binary description of a fork into a physical object?

Now, it happens that I have on my desk a lovely little object: a model of the three-dimensional projection of the four-dimensional analogue of a dodecahedron; a four dimensional object with 120 three-dimensional dodecahedra as its faces. (Okay, so I'm a math geek. Just look at it, it's cool.) They're made by my old friend Bathsheba Grossman, a mathematical sculptor, using a 3D printing process. Imagine, just for a second, how your regular ink-jet printer works. (If you don't have an ink-jet printer, imagine how my ink-jet printer works.) The print head scans across the paper, and the platen moves the paper upward; every place the paper should be black, the print head spits a tiny dot of waxy black ink at the paper. If you keep going back and forth over the paper, you can imagine how the waxy ink would build up to be thicker and thicker.

These metal objects are "printed" in much the same way, except to make them of metal, the "ink" is a kind of resin "glue" that sticks together fine metal powder. Later, the metal objects are heated, which drives off the glue and bonds the metal, a process called "sintering." Sheba then does a few manual finishing steps, but the result is an object that really can't be made by any other method --- and it comes off the printer in a matter of hours.

The ProMetal process Sheba uses is only one of several 3D printing technologies that use metals, plastics (video link), even waxes, and in almost all cases, right now, you get an expensive object in not very durable material. (Sheba's sculptures are the example that tests that rule: they're quite sturdy.) There are a few other areas where 3D printing and rapid prototyping is already useful, but they're pretty limited.

This is hardly the only process available, however. This notion of quickly turning a digital description into a physical object is becoming known as "rapid prototyping," with a range of possible technologies: numerically-controlled machining, various kinds of "3D printing" like Sheba uses, even numerically controlled woodworking. (You can find a nice summary of the different technologies at the University of Utah Rapid Prototyping Home Page, and this Rapid Prototyping tutorial page.)

We're pretty much at the MITS 8800 stage of manufacturing on demand -- there's no guessing which technologies will come out to be the winners. But we can make some guesses what the eventual results will be, if we just think about what the computer revolution has already done.

In all forms of publishing -- text, music, and video -- what the Internet (and to a lesser extent print-on-demand) has done is changed the economics. The up-front, fixed costs are very much reduced. You don't need to print a big press run to fill the bookstores; you don't need to press thousands of platters to move a record. What's more, the actual cost of producing the item is also reduced -- pennies for a book on my Kindle or a song on iTunes, or a dollar or so for a music CD. So now a musician can sell an album, or a writer can publish a book, without needing to fight for limited resources.

"Manufacturing on demand" -- or "fab," or "rapid prototyping" -- has the same potential. If you want a particular style of fork and can find it on the Internet, you can have it: send it to a fabrication company (or maybe even have your own at-home 3D printer -- you can already make a do-it-yourself version, like the [email protected] project ) and within hours or days you have your fork. If you have a design for a fork, you can sketch it with something like Sketch-Up, advertise it and sell it. No need to have a factory, no need to pitch it to a big company, no need to share the rights.

It isn't limited to small objects, either. There are new technologies, like those of the Contour Crafting project at UC Irvine, that can build houses using a similar, computer controlled technology. (What will it mean to architecture when you can design a house and then build the basic structure in a few days, automatically? In Chilé. Or on the Moon.) The Fab Lab at MIT uses more conventional technology, but then can build larger, more conventional things too. (What will car hobbyists be able to do if they can design and build a new engine on their desktop?)

What will it mean when everyone can be a designer and a manufacturer, just as computer technology lets everyone be a publisher, or cut and sell a record?

Start thinking about it. The time is coming.