A fundamental property of living systems is their ability to receive, process, and transmit information, allowing them to respond to environmental changes, search for food, and communicate. Uncovering physical mechanisms underlying such collective intelligent-like conduct in biological and synthetic systems constitutes an unsolved scientific challenge with a potential impact on understanding intelligent emergent behavior.
We will develop a theoretical framework and computational methods that will enable an understanding of the emergence of intelligent-like functions in populations of self-propelled units capable of exchanging and processing information transmitted by wave signals. In response to incoming waves, these units relay these signals by synchronizing their intrinsic states. We will develop computational models for communicating simple self-propelled units that are capable of executing preprogrammed tasks, such as self-healing, collective threat detection, and coordinated migration or dispersal.
We hypothesize that these information processing systems will enable elevated robustness to fluctuations, heterogeneity, or “adversary” interference. We will explore an “ecosystem” of agents that differ in their types of motility and their ability to sense and respond to acoustic signals: non-motile, facultative motile, or obligate motile, akin to categorizations found in bacterial colonies. The agent diversity will facilitate the richness of new functional intelligent-like behaviors.
We will apply machine learning, genetic and evolutionary algorithms, game theory, coarse-graining, and statistical mechanics methods to optimize the characteristics of emergent states. The project will stimulate the development of a self-sufficient, intelligent platform for swarms of synchronizing micro-robotic units.
