« Genetic Backup Files | Main | Mandatory Mileage Gauges? »


reprappump.jpgLast week, Alex wrote about the Future that Fabbing Suggests. Some of what he talked about is still some years off -- but some is a bit closer than you might think.

RepRap is a design for a Replicating Rapid-Prototyper ("rapid prototyper" being a more common industry term for fabber). While it hasn't yet been built, nothing in the design is outside the realm of what's currently possible. The designer, Dr. Adrian Bowyer of the University of Bath's Centre for Biomimetic and Natural Technology, sees it as a way to kick start the fabbing revolution. Bowyer intends to put the entire design, when completed, on the web for free download under the GPL -- he's already made the software for a RP-built robot available. The RepRap would work in a manner similar to ink-jet printers (its syringe pump is shown here), and would be able to embed conductive wires within plastic shapes. The RepRap would not technically be self-replicating, as it would only be able to produce its own parts -- some assembly would be required.

What's really interesting, though, is the question of what happens if RepRap works.

Bowyer envisions RepRaps as eventually costing about $500. What happens when such relatively low-cost 3D printers become available? For every "napster fabbing" scenario, there are sober rapid prototyping specialists arguing that large-scale mass production will still be more efficient than individual home factories. The latter will likely be true, at least initially, as long as the intent is to try to reproduce in one's home exactly what one could buy in a store. It's a less compelling argument, however, if RepRaps are used to make objects with a degree of customization and individuation unsupportable within a mass-production model.

There are a couple of ways that scenario could play out. One is akin to free/open source software: the products are so complex that only specialists design them, although quite a few are self-taught; good designers and functional components are well-regarded, and tend to show up in multiple projects; most people who use the designs appreciate the open, collaborative heritage, but end up being simply consumers, not contributors. The other is in a manner closer to website design: the first tools are sufficiently opaque that only the dedicated work with them; the next set make design easy enough for anyone to do, and the vast majority of the results are painfully bad; finally, the tools themselves do so much hand-holding that few really horrible designs result, but there's a level of uniformity approaching that of mass production anyway. Which path is taken depends on the complexity of the process.

Of course, it's possible that the actual path will be more of the first blending into the second, as the technology becomes easier to use and more widely available.

RepRap will likely be crude, clumsy, more expensive than intended, and ultimately a bust -- but, if even marginally successful, it will blaze the trail for better, more efficient, more cost-effective designs which can learn from RepRap's mistakes. Such follow-on designs could even emerge as variations of RepRap itself. Bowyer is making the right decision in putting RepRap's design on the web as open source -- he may not end up with the perfect design, but with enough hands and minds on the problem, someone eventually will.


This page contains a single entry from the blog posted on March 23, 2005 3:16 PM.

The previous post in this blog was Genetic Backup Files.

The next post in this blog is Mandatory Mileage Gauges?.

Many more can be found on the main index page or by looking through the archives.

Powered by
Movable Type 3.34