(ORDO NEWS) — Circling is observed in large groups of animals at different stages of evolution, from insects and worms to fish: they move in concert around the center of the group. The biological function of such bizarre behavior has so far baffled evolutionary biologists and systems theorists.
The standard approach to explaining circling postulates the presence of some artificial forces that arise between animals and affect their collective movement. Scientists at Skoltech took a different path: their model is goal-based and formulated in terms of reinforcement learning, a tool from the field of artificial intelligence.
Based on simple rules and natural constraints, the animals modeled by the researchers through trial and error learned to move collectively. Strikingly, the demands placed on them – to stay within the given limits from the center of the group and from the neighbors – led to spontaneous circling.
Even more unexpectedly, circling turned out to be very useful for survival: according to researchers, a team trained in such behavior is hundreds of times more resistant to harmful environmental disturbances – in nature, they can be, for example, wind or underwater currents.
Another application of artificial intelligence in a similar context is grouping animals. Birds migrate, wolves hunt, and fish swim in packs. If animals move together in an optimal position relative to each other, they thereby minimize energy costs.
The Skoltech researchers used the same goal-oriented approach based on reinforcement learning and confirmed that as a result, animals find the optimal configuration for movement, allowing them to save strength: two of them line up, three of them form a triangle, four of them form a diamond.
These and other, sometimes unexpected constructions, with a larger group size, were independently found by an alternative method.
Although there are other factors in nature that determine the structure of the pack, such as protection from predators, the position of the leader, and so on, the results obtained clearly demonstrate the wide applicability of the method and its reliability.
As Professor Brilliantov says, “I understand that everything is built from simple ‘mathematical building blocks’, but I never cease to be amazed at the possibilities of AI methods!”.
Contact us: [email protected]