Active gyrotactic stability of microswimmers using hydromechanical signals

Abstract

Many plankton species undergo daily vertical migration to large depths in the turbulent ocean. To do this efficiently, the plankton can use a gyrotactic mechanism, aligning them with gravity to swim downwards, or against gravity to swim upwards. Many species show passive mechanisms for gyrotactic stability. For example, bottom-heavy plankton tend to align upwards. This is efficient for upward migration in quiescent flows, but it is often sensitive to turbulence which upsets the alignment. Here we suggest a simple, robust active mechanism for gyrotactic stability, which is only lightly affected by turbulence and allows alignment both along and against gravity. We use a model for a plankton that swims with a constant speed and can actively steer in response to hydrodynamic signals encountered in simulations of a turbulent flow. Using reinforcement learning, we identify the optimal steering strategy. By using its setae to sense its settling velocity transversal to its swimming direction, the swimmer can deduce information about the direction of gravity, allowing it to actively align upwards. The mechanism leads to a rate of upward migration in a turbulent flow that is of the same order as in quiescent flows, unless the turbulence is very vigorous. In contrast, passive swimmers show much smaller upward velocity in turbulence. Settling may even cause them to migrate downwards in vigorous turbulence.

Reinforcement learning for pursuit and evasion of microswimmers at low Reynolds number

“…< l a t e x i t s h a 1 _ b a s e 6 4 = " 8 q Learning (RL) approaches rely on trials and errors to improve the quality of the decisions made by an agent -here a microswimmer. The potential of RL for navigation in complex and dynamic fluid environments has been demonstrated in various tasks, both in silico [14][15][16][17][18][19][20][21] and experimentally [22,23].…”

Surfing on Turbulence: A Strategy for Planktonic Navigation

“…Problems of planktonic navigation have been recently approached using reinforcement learning [20][21][22][23][24][25]. These studies showed that strategies based on local gradients can be learned in simple flows.…”

Surfing on turbulence

“…Problems of planktonic navigation have been recently approached using reinforcement learning [19][20][21][22][23][24]. These studies showed that strategies based on local gradients can be learned in simple flows.…”

“…These studies showed that strategies based on local gradients can be learned in simple flows. Training a microswimmer in unsteady 3D turbulence remains however challenging [22,23]. Moreover, the strategies learned are not necessarily optimal nor easily interpretable.…”

“…So far, we have assumed that the plankters can instantaneously choose a swimming direction p. This hypothesis is relevant for some copepods species that respond very fast to flow sensing compared to the characteristic Kolmogorov time scale [38], but a finite time is generally needed for plankters to reorient. To study the effect of such a delay, we introduce a reorientation time τ align [33], which results from the ratio between the viscous torque and the torque applied by the plankter (either passively by bottom-heavy plankters [39] or actively by surfers [19][20][21][22]). The surfers and risers considered so far correspond to the case τ align = 0.…”