Nature https://doi.org/10.1038/s41586-019-1022-9 (2019)

Emergent behaviours are arguably one of biology’s better tricks, and robotics would be well served to follow suit. Now, Shuguang Li and colleagues have implemented a robot comprising particles that move as an amorphous collective — remaining robust in the event of partial malfunction.

The system comprises ring-shaped particles undergoing cycles of expansion and contraction with a phase that is modulated according to each particle’s position. Li and colleagues induced autonomous locomotion by means of an algorithm familiar from collective cell migration: setting the phase offset proportional to the intensity of an external signal circumvents the need for interparticle coordination.

The team fabricated small robots to demonstrate their locomotive ability, as well as their capacity for moving and avoiding other objects. And unlike many robots, in which failure of a single part can result in malfunction, simulations of much larger systems showed these robots were robust to failure — moving at half their maximum speed even when 20% of the particles ceased to function.