Scientists at the Center for Bits and Atoms at the Massachusetts Institute of Technology (MITfor its acronym in English) are working on the development of robots that can build themselves.
The study, published in Nature Communications Engineering and quoted on the website of Urban Technoreveals the advances on these machines, capable of assembling any structure efficiently and affordably, no matter how large it may be.
These robots would be able to Assemble even bigger robots, build race cars, or build airplane wings. For each of these purposes, researchers have shown that assembler bots and subunits can be used, in large numbers, capable of independent movement.
As if it seems little, the robots will work autonomously and could build themselves, although for this last point they need more years of tests and studies.
The importance of artificial intelligence
Neil Gershenfeld, a professor at MIT’s Center for Bits and Atoms, explained: “When you build these structures, you have to build intelligence within them. The idea that emerged was the use of structural electronics, to fabricate voxels (the cubic unit that makes up a three-dimensional object) that can transmit power, data, and force. There are no cables, there is only the structure”.
Indeed, test robots are made up of voxels, which can be used to improve their structure and the part in which artificial intelligence enters. the new system allows machines to be aware of this limitation and have the ability to decide that it is time to build a larger version of themselves.
Aaron Becker, another of the MIT experts, assures that the possibilities are limitless: “This study examines a critical area of reconfigurable systems: how quickly the robotic workforce can be scaled up and used efficiently to assemble materials into a structure.”.
“This is the first work I’ve seen that attacks the problem from a radically new perspective, using robot parts to build a set of robots whose sizes are optimized to build a given structure (and other robots) as quickly as possible.”, he concludes.