The Fabulous Fab Lab

Researchers at Carnegie Mellon compete with those at MIT on a yearly basis. However, lately MIT has also been aspiring to a higher ideal: Star Trek.

MIT's Center for Bits and Atoms (CBA) is not composed of dreamers, though. Staff and graduate students have been working over several iterations to produce what they are dubbing "fab labs" -- or fabrication laboratories -- and are installing them in a mesh across the globe, from above the arctic circle in Norway to the crowded central region of India. Their latest installation, in Boston's sister city of Secondi-Takoradi in Ghana, was just installed this summer. Under the directorship of MIT Professor Neil Gershenfeld, CBA has created a machine that hopes to advance the very nature of manufacturing.

"It's just one machine with a great user interface," explains graduate student Amy Sun, who helped install the Ghana lab. While she says that current technology doesn't allow the labs to create anything useful on an atom-by-atom basis, as they do in Star Trek, right now they'll "settle for making nearly everything out of raw resources."

This includes everything from devices that measure water quality and engine efficiency in rural India to wireless network components and animal tags for reindeer herders in Norway. "An eight-year-old can create and assemble a fab lab oscilloscope," says Sun. The key to each lab is its ability to be highly customizable. When asked what the labs could make, she replies: "this is one of those really annoyingly difficult questions to answer, because the limitations really aren't in the machines but in your ability to imagine a solution."

Each fab lab has benefited from previous iterations. "At the moment," Sun says, "it integrates open-source software with open-source hardware" -- meaning commercially available components and instrumentation. The latest lab, in Ghana, includes three PCs (Knoppix and Debian Linux), a laser cutter, a combination desktop mill and 3D scanner, and a desktop-sized numerically-controlled Exacto knife. The machines are precise to 1/1000 of an inch, which Gershenfeld describes as "access to microns and microseconds;" and final products aren't limited to the bed size of each machine, either. Sun explains that "really elegant 3D structures can be assembled from smaller 2D geometries. Ikea is good proof."

Right now there are two limitations that the labs work against, she says. Physically, most of the labs are not able to work with large pieces of metal. Both the mill bed and the laser cutter are too small -- at 9" by 6" and 24" by 12" -- and the laser cutter only works with burnable materials. However, MIT is currently working with a waterjet cutter which operates similarly to the laser cutter: the difference, Sun says, is that "the manufacturer tells us that it [the waterjet cutter] will happily cut through up to six inches of hardened tool steel. Even a lesser version of this machine would be very enriching in the field."

The second limitation that Sun sees is knowledge sharing across the seven fab labs currently installed around the globe. Language and cultural differences also make coordination difficult, as well as other, more localized problems that she witnessed in Ghana: "misuse, hoarding, over-protectiveness." However, she is optimistic about the project's future.

"I think it is precisely these differences that we are interested in, because in addition to chasing after a machine that makes anything, we want a universal machine that makes anything. What's just so awesome about this experience is that even without the final product, just the approximation of the capabilities is already showing global impact. What happens when you engage the billions of brains all around the world and equip them with the tools to come up with technical solutions to local problems?"

If all goes to plan, an eighth lab will be installed in Washington, D.C., this fall, and other international sites are being considered as well. These are stepping stones to the goal that Sun and her colleagues have set for themselves: in a decade, the labs should be able to even replicate themselves.