RepLab: Machines Making Machines.

Look around you. Look down, past your hands, at the keyboard beneath. Look at the frame of the screen you’re reading this by. You are seeing the products of machines.

Chances are you’re surrounded by them, from the light over your head to the pipes and beams below. You are clothed by the products of machines. Kept warm by them. Transported by them. They cultivate your food, keep it cool, and help you cook it.

Where are these machines? Have you met them? Do you know the humans who tend them? What are their lives like? What would happen to you if they stopped tending those machines, or if the machines stopped running because the oil was gone?

Where would you get shoes, once the old tires were gone? How could you keep your rags together once you broke your last needle? You know windmills exist, boilers and steam engines and mirrors. Could you build them, and provide power for everything we take for granted?

You could if you had the right machines. And if they were just the right machines, they could make you most of what you need and a whole lot of what you want. They could even make more of the right machines…

RepLab is a project to build the right machines for the job. Serious industrial tools that can shape metal, assemble circuit boards, print in plastic, ceramic, and glass, and otherwise build most of what we need for high-tech living from scrap metal and specialized components like computer chips. All open source, and human scale: nothing you’d need a mortgage to build. Actually, all you’d need is some sweat equity, scrap metal, specialized electronic components, and a RepLab.

RepLab, as the portmanteau would suggest, draws from both the FabLab idea of a laboratory where you can build anything, and the RepRap project’s dream of a self-replicating desktop factory. We aim to design and build a suite of machines for general-purpose production of designs in most of the materials we take for granted: metal, plastics, wood, ceramics, and ultimately glass and thread as well.

The project is practical in emphasis. The call to action for RepRap was penned by Marcin Jacubowski, a visionary builder working to create resilient, high-technology community at the agrarian village level. His concept of permafacture combines the agricultural sustainability of permaculture with a high-tech open source ecology of technologies which can be replicated on a scale of about a hundred people, anywhere in the world. His lab and home, Factor e Farm, has produced impressive results, including an open-source compressed earth block press, an open-source modular tractor for cultivation and construction, and a computer controlled table that cuts steel with plasma.

The last project, called RepTab, is being adopted and modified as one of the first RepLab projects. The table is made mostly of angle steel, which can be cut and drilled with the plasma torch head of a RepTab. The RepTab can make most of its structural pieces, automatically. That’s a recursive machine, and it’s the very beginning of what RepLab is all about.

We’re trying to be fairly precise about our goals, so that we’ll know when we’ve achieved them. For example, we avoid the term ‘self-replication’ in reference to our machines. Self-replication, as practiced by cells, includes self-assembly. This is a hard problem, and not a project goal.

We speak instead of recursive manufacture, and general replicators. Recursive manufacture is simple: it is building a widget using the widget. It is important, but not particularly impressive in and of itself. An example is a kiln, which is simply a pile of bricks and a fire. Since bricks are made in a kiln, and trees grow themselves, a kiln next to a forest and a vein of clay is a fully recursive technology. It is also manual, arduous, time-consuming work.

Recursive manufacture is the rule, not the exception. Everything is made on machines, including the machines that make things. In fact, the traditional factory is split between the production floor, where machines make products, and the machine shop, where machines make tools for making products.

What has changed is the nature of those machines. Consider the single case of writing. To replicate a text, a scribe needed tools: a reed pen, ink, a grinding stone, and papyrus, a knife to trim the reed. Without the tools, the job is impossible. With the tools, the job can be done.

One thing to notice is that the human is working, and the tools are helping. A faster tool, a pen rather than a chisel, makes faster work, as does practice. But all work is human work, and it shows: in particular, skill becomes an all-important factor in the quality of the work and how much can be performed. This kind of replication scales poorly, favors one-of-a-kind items, and rewards lifetime specialization. Call it craft; we call it Type 1 replication.

The printing press was one of the most disruptive technologies of all time, and the reason is that it was one of the first examples of a new kind of replication. Now, by investing significant time in building a machine, casting type, and laying out a page, a human can replicate a text hundreds or thousands of times, just by pulling a lever. Now we have a technology for replication that favors mass-production, scales well, and allows for the exploitation of unskilled labor. This is Type 2 replication, characteristic of the Industrial Revolution.

Type 2 replication results in an explosion of basically identical, cheap products, a lot of jobs in the cities, and a social and industrial ecology dominated by capitalists, who have the money to enjoy the benefits of large-scale manufacture. It remains the dominant form of manufacture, and it shapes our way of life, right down to the concepts of 9-to-5 and the weekend.

The Type 2 tools have evolved in a couple directions, which converge on our goal, the Type 3 replicator. As printing presses evolved, they became fully automatic in operation. A large early 20th century newspaper press still had to be laid out; once this was done, the machine was turned on, the paper rolls started spinning while the ink vats emptied, and bound stacks of newspaper shot out the other end. This is Type 2a replication, ‘a’ for ‘automation’. It displaces Type 2 wherever it is developed, since one no longer has to pay workers to pull levers.

The other direction is exemplified by the early Xerox machine. It could make a copy of any template provided, quickly and automatically, but needed human tending to change the template or do anything other than spit out identical copies, such as collation. For anything but a single sheet, an old-fashioned Xerox is a Type 2b replicator, ‘b’ for ‘bespoke’ or ’boutique’. They favor customization, and scale poorly, but they give good results to unskilled operators. Type 2a replication gave us the magazine; Type 2b replication gave us the ‘zine, which were culturally important without ever displacing the more professional media that inspired them.

The breakthrough to Type 3 came with the laser printer. A laser printer can print whatever template you want, accurately and quickly. It is functionally just as easy to make 1000 different pages than to make 1000 of the same page. The largest laser printers take a variety of raw goods in one end, and out the other end comes books, printed, cut, collated, bound, wrapped in a book jacket, packed in a mailing envelope and stamped. Each book different, each book ready to go. Humans write the books, order the books, empty the hoppers, and refill the feedstocks. The rest is all machine.

Type 3 replicators, in a variety of media, are the goal of the RepLab project. They scale well, favor customization and human empowerment over mass production and serfdom, and tend to economically displace Type 2 technologies when they emerge.

The computer is the best Type 3 replicator ever invented, in fact I would argue it is impossible to have a Type 3 replicator that is not also a computer. It is the computer, not the laser printer, which is displacing the magazine and the newspaper, through its ability to replicate any text in any language for the cost of a sip of electricity. Vinyl records are a Type 3 replicator with a Type 2 production process for the templates. Mp3 players are Type 3 through and through, and that’s the single biggest difference.

So lets get down to brass tacks, which we hope to eventually print with brass powder and a laser. The RepLab project has big ambitions and an open-ended project list, but we’ve already whittled down to a few promising and existing technologies to adapt and pursue.

The first one is a RepRap type printer, with similar function and different design goals. The RepRap aims to make a small printer for 3-d plastic objects, which can make as much of its chassis as possible. Our printer will be based on the RepRap extrusion platform and source code, but modified to be robust and automatic, so it can print plastic parts, one after another, for weeks at a time, with someone stopping by from time to time to empty the hopper and feed it another roll of plastic filament. Recursive manufacture is a goal for the whole RepLab, not each of the machines in it; we will use printed plastic where it makes sense in the design, and other approaches elsewhere.

The second one is based on the plasma cutting RepTab described before, but modified to be capable, with modular improvements, of routing, plasma cutting, and possibly laser and water jet cutting. Add a feed for 4×8 sheets of metal, wood and foam, and an automatic hold-down, and you have a robot that can make almost anything that’s flat, with a minimum of adult supervision.

The third goal is to combine the work of the CandyFab project with that of Open3dp to produce an open-source powder-bed printer. With some development, this machine will be able to make ceramic and opaque glass objects and bind metal powder for heat sintering. Combined with a simple furnace/kiln, and there are shelves full of free designs for those, this offers the promise of a RepLab that can print everything AND the kitchen sink! Plus molds for metal casting, thermal insulators, and many other useful and wonderful things.

There’s more we’ll need, and there’s more we’re planning to build, in particular a system for automated circuit assembly. Even these three technologies, in tandem, would provide a facility which could produce many economically significant goods, including a fair capacity for recursive manufacture.

The project is in the very early stages. We’re building a crowdfunding platform and working on moving forward with our core designs so we can build them as rapidly as practical. To join the conversation, stop by our google group, follow us on Twitter or at @RepLab, and look for our project blog coming soon at

Better yet, get some friends, get a space for a RepLab, and start hacking! We intend to build and replicate these technologies all over the world. We hope you’ll join us.

Published in: on November 28, 2009 at 7:41 pm  Comments (3)  

The URI to TrackBack this entry is:

RSS feed for comments on this post.

3 CommentsLeave a comment

  1. […] Via Sam Putnam: […]

  2. Nice description and framework for the whole problem.

    Please, don’t forget to look into A.M. Fogassa’s work @ He has the plans for a fully functional 3DP system which is a DIY.

  3. And of Course, I would be remiss in my duties if I didn’t direct all the readers of this post to where the struggle to crowdsource continues, to build what could be a core RepLab system

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: