If Terminator 2 taught us anything, it was that a properly timed thumbs-up can make us cry. Also, autonomous, self-shaping blobs are a must-have on our checklist for the future.
Now, MIT professor Daniela Rus and student Kyle Gilpin are publishing a paper on building such a wonder goop. They tend to refer to the technology as a smart "sand," and they imagine scenario in which you could drop a small model into a vat of the sand, and the sand could sense that model’s contours and create a 3-D version of the object from that information. So you could drop in a tiny cup, the sand would sense the cup’s negative space and then it would shape into a cup that was 10x (or much more) larger.
Right now, the team is focused on developing an algorithm to make their approach possible—a hyper-efficient language that’s simple enough for each grain of sand to understand without massive processing power. It’s only with this language that the idea could hope to scale.
"The beauty of encasing a prototype of the object to be formed with the smart material is that we drastically reduce the communication burden on the system so that we are prepared to scale-up the number of modules in the system," Gilpin tells Co.Design. "If we used CAD or similar, we’d be forced to transmit a complete description of the shape to be formed to all of the modules. With 100 modules that’s okay, but somewhere on the way to a million, it becomes unreasonable. We don’t want to send a million messages, one for each module in the system, telling each whether it is a part of the shape we’re attempting to form or not."
To test their math, the teams has developed a prototype of the sand. (Gilpin calls these larger pieces "pebbles.") Each pebble is 10mm across and contains an independent processor along with magnets that enable the magical sticking trick. As cool as they may be, the resolution of this rapid manufacturing technology is really only as sharp as the building blocks are small. Think of them as 3-D pixels.
"Shrinking the hardware presents the biggest long-term challenge," admits Gilpin. "As we shrink the modules, we’ll have to look for alternative connection mechanisms. One possibility is using electrostatic forces instead of magnetic ones. This would allow us to replace the relatively bulky electro-permanent magnets with much smaller electrodes."
That said, the technology’s promise is massive. You could build structures that were far smarter than even our most advanced 3-D printed parts (even those that we’re using within the human body), that can respond and adapt to their environment. "In addition to duplication, I would see our modules used for sensing tasks in constrained spaces," writes Gilpin. "Perhaps you could pour our modules down a pipe in order to both map the shape of the pipe while finding defects or cracks. The modules could potentially help reinforce those weak points…Because it can self-disassemble, perhaps it could be used as an intelligent scaffolding for bone, or even organ, regrowth. The system could sense and relay important information to doctors and then disassemble as the bone healed."
Gilpin knows that the technology is a ways off, believing it to be closer to a 10-year vision than a five-year one. In other words, this self-assembling sand could be waiting for us right where 3-D printing leaves off.
[Images by M. Scott Brauer/MIT]