Research into self-replicating rapid prototyping machines could create a production singularity, after which free objects, communication and medicine could be the norm.
The Industrial Age centralised the production, increased the quantity, and homogenized the design of manufactured goods. Traditional arts and crafts techniques dwindled to niche markets, and factories creating hundreds of thousands of identical items a day became the norm. Products became significantly more complex, and now it would be impossible for consumers to make most common household items themselves.
However, a paradigm shift may be on the way in the form of cheap rapid prototyping machines for the home. Cheap because you don’t have to buy one, your neighbor can ‘breed’ you one via self-replication.
The designers behind these machines have a formidable but admirable goal; to democratise the production of household goods. The Rep-Rap project at Bath University vocalises this goal with its slogan, “Wealth without money.”
Some recent advances have already allowed the enthusiastic hardware-hacker to create certain items at home; it is possible to make custom-designed PCBs using a laser printer for example. But these have required some technical know-how on the part of the user, and a vast amount of tinkering. Normal rapid prototyping machines also allow production of certain parts, albeit with an enormous price tag and large space and power requirements.
The RepRap project at Bath University, headed by Adrian Bowyer, aims to create a cheap rapid prototyping machine that also allows the creation of printed circuits. Their method of lowering the price is ingenious â€ enable the machine to self-replicate. One person with such a device could make two more and distribute them. The two recipients could also make two more, and so on. This exponential production could create millions of these machines very quickly.
â€œWhat we are attempting to do is make a machine that can produce almost all of its component parts, to make self-replicating machines.â€ says Bowyer.
To do this requires creating a machine that walks a fine line between complexity and simplicity; complex enough to manufacture its own constituent parts, but with constituent parts simple enough to allow this.
The team arenâ€™t far from their goal according to Bowyer: â€œWeâ€™re at the stage now were we have almost all of the component pieces of the machine designed and tested, and what weâ€™re interested in now is tying them all together.â€
In order to improve the design, and to allow replication, the plans will be made freely available under the GNU licence, as will the software to design parts and control the machine.
The project certainly isnâ€™t lacking in support. A flurry of press attention in 2005 brought a number of people to the project from all over the World. Currently there are teams in New Zealand, the US and Canada, and theyâ€™re all working on a volunteer basis. Vic Oliver, a team member in New Zealand, has very recently got a small machine working, which produces small components, such as small plastic polo-mint shaped pieces. The American team are about halfway to finishing a working model. These teams are all coordinated and supported by Bowyer and his team in Bath. At their lab they have the luxury of a rapid prototyping machine, supplied by a government grant, and they use it to manufacture parts for other machines as well as their own. The next goal, once these distributed machines are working, is to combine the best elements of each to produce a â€˜Version 1.0â€™ machine.
Currently, some parts to create a machine will be bought in, rather than produced by a parent machine as it makes economic and practical sense, says Bowyer, â€œWe buy in nuts and bolts as it is cheaper and easier. It would be nice to have a pure machine, but I would imagine that it would be some time before these readily available parts are produced by the machine itself, even when they are possible to be made. It will be a long time before people bother because theyâ€™re available across the whole planet, very cheaply. Itâ€™s a practical problem rather than a philosophical one.â€œ
The ultimate aim is to create a machine that can also create electrical conductors, and once that ability is combined with the 3D plastic forming any device can be created, even electrical motors. According to Bowyer this could happen within 2 years.
â€œTo get this conductor extruded is very easy there are two alloys, woodâ€™s metal and fieldâ€™s metal, which have lower melting points than that of the plastic used in the machines construction. Fieldâ€™s metal would be preferable as Woodâ€™s metal contains lead. We would have a small chamber that heats this alloy and we can then extrude this from the same nozzle that we would use to extrude the plastic. The plastic component would have a channel left there in the design, which we fill with the alloy.. The big advantage is that you can then put a layer of plastic on, and another layer of alloy â€ you have 3D circuits. You can then hide these circuits in the actual structure of the machine, in the frame of the machine.â€
Once these machines are capable of reproducing, and once some spare parts have been formed and safely stored, then they could spread exponentially.
It is then that they can become productive, and could be used to create any household product. It is envisaged that a vast library of open-source designs would be made available online, created and refined by users, just as with open source software.
Subsequent machines with refined designs will be able to create ever more complex components and devices, explains Bowyer, â€œOnce we have these simple things right thereâ€™s a very clear path to making complex things. Making complicated things is not a large step, certainly not as large as the step from making nothing to making simple things.â€
There are some potentially revolutionary ideas that arise from the ability to create complex devices at home, most notably in communication and medicine.
An open source, peer to peer, mobile phone network that doesnâ€™t rely on a service provider would be made possible by creating phones that also act as base stations, routing traffic from one phone to another towards it’s destination. The absence of a service provider would theoretically make the service free. Free mobile communication could have a phenomenal impact on the whole world, but developing countries in particular would benefit.
Another exciting area that reprap is anticipated to make an impact in is medicine. The idea would rely on the fact that patents donâ€™t apply to personal production. If an individual has the means to make a copyrighted drug, then they are freely allowed to do so as long as it is for strictly personal use. The only problem is that no one does have the means. Bowyer imagines that reprap could change this.
It is almost cost effective even now, to produce drugs at home in certain scenarios, as Bowyer explains: â€œTake Herceptin for example. A yearâ€™s dose costs Â£20k, thatâ€™s a lot of money. When you look at home drug replication machines, they cost Â£70k-100k and thatâ€™s more than Â£20k, but not a lot more when you consider that someone may be on a drug for many years.â€
Bowyer believes that insurance companies, particularly in the US, will begin to take advantage of this. Reprap machines could take this democratization even further, enabling people to create drug replication machines cheaply within their own homes, and reducing the cost of even the most expensive drugs to little more that the price of its raw materials. â€œOnce that idea takes off you have a complete change in the pharmaceutical industry, and thereâ€™s no reason why the rep-rap machines couldnâ€™t produce these drugs, or machines to produce them. We need a wide uptake to get these machines out there, to make them popular. Back in the early days of computing I soldered together my first computer, and in a few years time that will be where reprap is: being made by geeks and hackers like me.â€
The supply of raw material is as bizarre and inventive as the rest of the project. Initially it is envisaged that they will be supplied commercially, but in order for this machine to realize its full potential the cost needs to be kept to a minimum, and the availability needs to be universal. Bowyer has been investigating a polymer called Poly-Lactic acid that can be created by fermenting starch, either from corn or potatoes. By growing the raw material you have an almost carbon neutral source of polymer, which is local to the machine. In the developing world this will be very vital.
The whole point of this machine is that it reproduces, so that you have something that is analogous to a biological system, and because it reproduces it will evolve. People will always be changing the design, to improve and diversify. However there are some core aspects to the machine that will probably be widespread, for example the software.
The software is currently being written in Java, and the machines will require a computer in order to control them. The machine will probably connect via the USB port of the computer, and because of the platform independence of Java, can run on a multitude of computers. One particular project that the team have been watching closely is MITâ€™s $100 computer, as it would provide a perfect controller for the device. It is designed to run from 12 volts, just as the reprap machine, so could use the same power source, and is capable of running Java code.
The cheapest rapid prototyping machine currently available still costs Â£12,000, whereas the parts and materials required to produce a reprap machine will cost Â£300. The price is important, because itâ€™s within the boundary of what people will pay for a home appliance. Itâ€™s really no different to other rapid prototypers, except that every component part can be made by the machine itself.
The benefit of these machines is yet to be seen, but a tipping point is incredibly close. Perhaps even by the time this article is published the first machine will be finished, which will be capable of producing most of its own parts. Once this has happened there will be a rapid surge in the machineâ€™s capabilities. Constant design refinement will enable more complex parts to be produced, which will enable more complex design refinements. Obviously there is a limit to this refinement, but does that limit lie above or below the level needed to produce a drug replication machine, or a peer to peer communications infrastructure?