We’ve seen a few different robotic construction projects that leverage stochasticity to assemble themselves into complex, ordered shapes or structures through a combination of time and random motion. Generally, the autonomous protorobot modules (is that a thing?) are active on some kind of artificial random motion generator like a table that shakes, but it’s much more fun to see autonomous modules assembling themselves using only the power of nature.
This project of helium balloons carrying small structural segments is called Aerial Assemblies, and it comes from MIT researcher Skylar Tibbits and MIT’s Self-Assembly Lab:
Aerial Assemblies focuses on large-scale modules that can assemble in the airspace high above land, construction sites or complex environments. After the helium fades and the modules touch-down, the assembled lightweight structural lattices will remain.
Which shapes the modules create depends on which of their connection points are activated, using magnets or whatever mechanisms are appropriate for the final structure. The assembly process itself is completely autonomous, depending only on enough energy being input into the system to keep the modules moving randomly relative to each other.
In this particular case, that just means a windy day, although the same assembly principle can be applied to modules in turbulent water, and the Self-Assembly lab has managed to create a chair in this way:
The overall idea here is that eventually it may be possible to use techniques like these to create very large, complex structures in places that wouldn’t otherwise have the space or infrastructure necessary to do so, like in space, in the developing world, in any disaster area, or on a scale too tiny for anything but self-assembly. As Tibbits explained to Vice:
One long-term goal of these experiments is to develop a construction process that needs neither machine nor human intervention, burns no fossil fuels, requires no extra energy, can run 24/7, and is optimal for situations that pose a lot of problems for existing technologies.