The energy required to initiate self-assembly in the MIT/Stratasys project comes from interactions of the water molecules with the molecules of the water-expanding material, said Dikovksy. Other energy sources could include humidity, sound, heat, or vibration. But before that, the next step could be generating energy by removing water, which will make the structure contract instead of expand.
In an interview on the TED blog about his 2013 TED Talk, Tibbits says potential applications for the technology are space systems that expand and self-assemble in orbit, activated by changes in pressure, temperature, or light.
Self-assembly of artificial systems is not a new idea. It's being pursued at the nano-level, using carbon nanotubes and organic or engineered DNA, as well as various methods for modular, self-reconfigurable robots.
We've covered mechanical, self-assembling robots such as the Smart Pebbles robotic cubes built by a team in the Distributed Robotics Laboratory (DRL) of MIT's Computer Science and Artificial Intelligence Lab (CSAIL). At the nano-device level, we've reported on synthetic DNA strands programmed to self-assemble into 2D tiles, and more recently, into 3D bricks, by researchers at Harvard's Wyss Institute for Biologically Inspired Engineering.
Many of the developments in robotics are actually aimed at product manufacturing: The idea is to use robotic modules to make rapid prototypes, self-repairing systems, replacement parts for other systems, and self-reconfiguring systems like furniture that changes from a chair into a table. Adding expandable, programmable materials and 3D printing to this mix will give the development of this rapidly-changing field a big boost.