It’s been very cool to watch 3D printers and laser cutters evolve into fairly common tools over the last decade-ish, finding useful niches across research, industry, and even with hobbyists at home. Capable as these fabricators are, they tend to be good at just one specific thing: making shapes out of polymer. Which is great! But we have all kinds of other techniques for making things that are even more useful, like by adding computers and actuators and stuff like that. You just can’t do that with your 3D printer or laser cutter, because it just does its one thing—which is too bad.
At CHI this year, researchers from MIT CSAIL are presenting LaserFactory, an integrated fabrication system that turns any laser cutter into a device that can (with just a little bit of assistance) build you an entire drone at a level of finish that means when the fabricator is done, the drone can actually fly itself off of the print bed. Sign me up.
There are a couple different components that make up LaserFactory. First, you’ve got a commercial laser cutter to do what commercial laser cutters do— in this case, cutting a quadrotor frame out of plastic. Second, you’ve got a hardware-add on, which is a whole bunch of other stuff that reversibly bolts onto the head of the laser cutter. The add-on includes a silver paste dispenser, a little suction gripper to do pick-and-place, and some actuators, solenoids, and a small vacuum pump. Once the laser has cut out the base structure, the silver paste is dispensed wherever you’d need either a conductive circuit trace or something glued to something else. Then, the suction gripper adds components one by one, moving them from a preloaded storage area into the fabrication area. The last step is to use the laser once more to zap the silver paste to thermally cure it, turning the traces conductive and also soldering components together. Those conductive traces do look a bit messy (the paste spreads out after being deposited), but adding another step of engraving small channels into the substrate can help keep it contained to sub-millimeter widths.
The really clever thing that’s not necessarily obvious from the video is that the hardware add-on is not communicating directly with the laser cutter head that it’s attached to. You’d think this would be an absolutely necessary step, because otherwise how does the add-on know where its location is and what it should be doing? But it does know these things, just indirectly, by using an accelerometer to track the movement of the fabrication head, and then converting specific movements that the fabrication head makes into instructions. To ‘program’ the hardware add-on, then, you just have to embed some custom movement instructions into your fabrication program (like a very specific little shimmy), which the hardware add-on will then read in order to trigger a specific function. This method is a clever one because it makes the hardware add-on more or less agnostic to the kind of fabricator that it’s working with. As long as your fabricator accepts custom movement instructions, this motion-based signaling technique means that it can control the hardware add-on.
For more about what it would take to get this kind of thing working on a consumer 3D printer, and what the future may hold for personal fabricators, we spoke with first author Martin Nisser via email.
IEEE Spectrum: Besides laser cutters, what other fabricators could your system work with, and what modifications would be required?
Martin Nisser: Our motion-based signaling technique obviates the need to communicate with a particular fabrication platform by encoding the fabrication instructions into the fabrication file itself. For the laser cutter we used, this entailed embedding the instructions into a pdf. This feature makes our add-on agnostic to specific fabrication platforms so that it is portable not only between various laser cutter brands, but even 3D printers— by translating the fabrication instructions to G-code, we were able to deploy our system onto a 3D printer too.
The advantage of this is that our signaling technique can conceivably be used to deploy researchers’ custom hardware add-ons onto any machine that utilizes a 2-axis CNC platform, though modifications would be required first in hardware to physically connect it to different end effectors, and second in software to calibrate it to different accelerations. A drawback to this communications paradigm is that the communication is one-way: platform agnosticism means that the hardware add-on has no access to the fabrication file itself, and so relies on a form of dead reckoning which means it would be unable to self-correct in the event of an error.
What level of skill or experience does it take to operate LaserFactory? Who do you hope could benefit most from the kind of device that you're developing?
Once LaserFactory is assembled and mounted onto a laser cutter, it requires no further intervention to operate. In the near term, this kind of one-stop-shop for fabrication would be beneficial for researchers, educators, product developers and makers looking to rapidly prototype functional devices such as wearables, robots and printed electronics. It is also a compelling solution for logistically challenging environments such as space, where the ability to create functional devices remotely, on-demand and without human intervention is paramount.
More generally, users stand to benefit where they need to create physical prototypes for devices but may not have the skills to make them; people shouldn’t in the future be expected to have an engineering degree to build robots any more than they should have a computer science degree to install software.
In what ways could you potentially extend this system?
We hope to build on this technology by exploring how to create a fuller range of 3D geometries, perhaps by integrating traditional 3D printing into the process. In addition, we would like to chart the full design space of what we can make, for example by leveraging the full volumetric space of the laser cutter platform to create devices up to 1 [meter] or greater in length. Beyond the engineering, we are also thinking about how this kind of one-stop-shop for fabrication devices could be optimally integrated into today’s existing supply chains for manufacturing, and what challenges we may need to solve to allow for that to happen.
What kind of future do you envision for personal fabricators?
From Star Trek’s Replicators to Richie Rich’s wishing machine, the media has a long history of inspiring speculation about a machine capable of creating arbitrary “things” for people on demand. In research, the idea of a machine to make machines—such as von Neumann’s universal constructor— has also received serious attention, and researchers have in recent decades worked actively towards a long-term vision of being able to download a device file and have it fabricated at the push of a button. We hope that one day, fabricating custom hardware through personal fabrication machines will allow for as much personalization and be just as straight-forward as downloading software is today.