Against the white void, the assembled robots look like nothing so much as ball bearings with Christmas lights on.

This is a mass of kilobots, minimalistic robots that each contain an LED light, a vibrating motor and the ability to communicate with other nearby kilobots via infrared signals. And after a human tears a swath through their assembled clusters, the robots organically, intuitively, autonomously reform into a similar shape, guided by nothing so much as its own basic rules.

The experiment, by a team of researchers in Spain, the UK and the Netherlands, posits a simple question: Robotic swarms scan as intense, complex things, but what if people can engineer them to work together as simply as cells in organic matter?

To make the swarm work, the robots are assembled in a mass and given a flexible state. Following their coding, the robots learn how they are alike and different from the robots near them, then move to cluster with robots of a certain state, and then adjust to see if they are clustered as they should be.

In appearance, this means tiny blue robots vibrating to be closer to clusters of tiny green robots, with those on the periphery of the clusters forming a sort of natural teal halo. This works both for the robots left undisturbed and actively disturbed, the dense green clusters reforming like tissue stitching over a wound.

Robotic swarms are compelling for researchers, civilian and military alike, because they offer functionality through interchangeable components. Much of the promise of swarm robotics wants the machines to be as cheap and disposable as possible, while the swarm as a whole is still capable of performing tasks, even when components fail.

“Loss of robots can occur for a variety of reasons — loss of digital infrared (IR) messages, inaccuracy in distance estimation, and unreliable movements of robots — all of which relate to the general lack of reliability of these relatively simple, cheap kilobots,” the researchers write.

“Despite this, the swarm as a whole continued to control its global shape, even when reduced to less than one-third of its original size or when physically ‘damaged.’”

For now, research like this kilobot swarm assembly will remain in the lab, but the potential implications of cheap robots that can communicate and coordinate as effectively and simply as demonstrated here could be huge. Everything from structure reassembly to patterns of self-healing armor to autonomous scouting clusters could potentially be derived from studies like this.

Watch the robots in action below: