By Sebastian Anthony on August 8, 2014 at 11:00 am
Harvard & MIT create first self-assembling robots – the first real Transformers
Harvard and MIT engineers, showing a reckless disregard for the robocalypse, have created origami robots that can self-assemble themselves — from a flat piece of paper and polystyrene – and walk away in just four minutes. “Getting a robot to assemble itself autonomously and actually perform a function has been a milestone we’ve been chasing for many years,” says Harvard’s Robert Wood, barely stifling a maniacal cackle.
The origami robots are made from composite sheets of paper and polystyrene (Shrinky Dinks). The engineers print some conductive channels onto these sheets, and then use a laser machining system to create the necessary origami fold pattern. Each hinge contains an embedded circuit that, when instructed by a microcontroller, produces heat, which causes the hinge to fold. Then, by adding a couple of motors, and a microcontroller that knows the necessary fold pattern, the robot comes to life — first by folding into its predestined shape, and then by walking away. Watch the video: It’s really very cool. [Research paper: DOI: 10.1126/science.1252610 - "A method for building self-folding machines"]
While Harvard and MIT’s origami robots are a technological marvel, and rather cute to boot, it’s important for us to keep our eye on the prize — in this case, self-assembly. One of the greatest bottlenecks in technology (and society) is that, for the most part, we still have to assemble everything by hand. Robots and other automated processes can produce many of the individual components, but there are still millions of humans around the world who spend all day slotting widget A into sprocket B.
In theory, a self-assembling robot (or gadget) would know how to build itself from the individual components — much in the same way that almost everything in nature self-assembles from simple building blocks. As you can imagine, this would dramatically reduce the cost of production, and it would also enable some rather unique use-cases. It would be a lot more efficient if we could send flat-packed robots to the Moon or Mars, or drop a ream of self-assembling clean-up robots in disaster areas.
The caveat, of course, is that you’re teaching robots how to build themselves. Right now we’re just talking about dinky robots that can walk around — but how long do you think it’ll be before we create a self-assembling robotic production line? Yes, it would be absolutely awesome if you could simply drop a shipping container full of components and have it self-assemble into a factory — but what if the self-assembling robots become self-aware, start producing more factories, and then bypass their human orders and start producing robot soldiers that enslave the human race?
The Harvard researchers, of course, have a much more benign vision of the future. Basically, they imagine that you would have a 3D-printing shop that can churn out these self-assembling robots. “You would be able to come in, describe what you need in fairly basic terms, and come back an hour later to get your robotic helper,” says project lead Robert Wood. The current robots, before motors and batteries, only cost around $20 to make — so this is actually quite a feasible scenario. But then Don Ingber, another Harvard researcher goes and ruins it all by saying this: “The days of big, rigid, robots that sit in place and carry out the same repetitive task day in and out are fading fast.” I like my static robots, darnit! At least you always know where they’ll be…
web page
![[Linked Image from s2.excoboard.com]](http://s2.excoboard.com/forums/25521/user/302752/484538.gif)
