Harnessing Synthetic Active Particles for Computations

Synthetic active microparticles have complemented biological active matter with their self-propulsion capabilities, providing new propulsion mechanisms and a wide range of collective behaviors of motile active matter. The feedback control of active particles has endowed them with behaviors that even enable adaptive responses through machine learning [1]. However, the challenge remains to create functional synthetic active matter, i.e. active particle systems that perform functions such as information processing. Active particles do not have to be equipped with functions but should enable this function through their own dynamics.

We demonstrate an approach to information processing with active particles by using them for physical reservoir computing [2]. We show that synthetic active microparticles self-organize from an active microswimmer and a passive colloid into inherently noisy nonlinear dynamical units. The self-organization and dynamical response of the unit are the result of a delayed interaction of the microswimmer with a passive target. The individual units show due to a bifurcation a memory with relaxation times that depend on the activity of the microswimmer.

A reservoir of such units with a self-coupling via the delayed response can perform predictive tasks despite the strong noise resulting from the Brownian motion of the microswimmers. To achieve efficient noise suppression, we introduce a special architecture that uses historical reservoir states for output. Our results pave the way for the study of information processing in synthetic self-organized active particle systems and suggest that dynamical microscopic active systems can be used to implement new functionalities.