 | ||
Self-propelled particles (SPP), also referred to as self-driven particles, is a concept used by physicists to describe autonomous agents, which convert energy from the environment into directed or persistent motion. Everyday life examples of such agents are walking, swimming or flying animals. Other biological systems include bacteria, cells, algae and other micro-organisms. Generally, directed propulsion in biological systems is referred to as chemotaxis. One can also think of artificial systems such as robots or specifically designed particles such as swimming Janus colloids, nanomotors, walking grains, and others.
Contents
Overview
Self-propelled particles interact according to various social and physical rules, which can lead to the emergence of collective behaviours, such as flocking of birds, swarming of bugs, the formation of sheep herds, etc.
To understand the ubiquity of such phenomena, physicists have developed a number of self propelled particles models. These models predict that self propelled particles share certain properties at the group level, regardless of the type of animals in the swarm. It has become a challenge in theoretical physics to find minimal statistical models that capture these behaviours.
Biological systems
Most animals can be seen as SPP: they find energy in their food and exhibit various locomotion strategies, from flying to crawling. The most prominent examples are fish schools, birds flocks, sheep herds, human crowds. At a smaller scale, cells and bacteria can also be treated as SPP. These biological systems can propel themselves based on the presence of chemoattractants. At even smaller scale, molecular motors transform ATP energy into directional motion. Recent work has shown that enzyme molecules will also propel themselves. Further, it has been shown that they will preferentially move towards a region of higher substrate concentration, a phenomenon that has been developed into a purification technique to isolate live enzymes.
Artificial systems
One usually distinguishes between wet and dry systems. In the first case, the particles "swim" in a surrounding fluid; in the second case the particles "walk" on a substrate.
Active colloidal particles are the prototypical example of wet SPP: Janus particles are colloidal particles with two different sides, having different physical or chemical properties. This symmetry breaking allows, by properly tuning the environment (typically the surrounding solution), for the motion of the Janus particle. For instance, the two sides of the Janus particle can induce a local gradient of, temperature, electric field, or concentration of chemical species. This induces motion of the Janus particle along the gradient through, respectively, thermophoresis, electrophoresis or diffusiophoresis. Because the Janus particles consume energy from their environment (catalysis of chemical reactions, light absorption, etc), the resulting motion constitutes an irreversible process and the particles are out of equilibrium.
Walking grains are a typical realization of dry SPP: The grains are milli-metric disks sitting on a vertically vibrating plate, which serves as the source of energy and momentum. The disks have two different contacts ("feet") with the plate, a hard needle-like foot in the front and a large soft rubber foot in the back. When shaken, the disks move in a preferential direction defined by the polar (head-tail) symmetry of the contacts. This together with the vibrational noise result in a persistent random walk.
Typical collective behaviour
The prominent and most spectacular emergent large scale behaviour observed in assemblies of SPP is directed collective motion. In that case all particles move in the same direction. On top of that spatial structures can emerge such as bands, vortices, asters, moving clusters.
Another class of large scale behaviour, which does not imply directed motion is either the spontaneous formation of clusters or the separation in a gas-like and a liquid-like phase, an unexpected phenomenon when the SPP have purely repulsive interaction. This phase separation has been called Motility Induced Phase Separation (MIPS).
Examples of modelling
The modeling of SPP was introduced in 1995 by Tamas Vicsek et al. as a special case of the Boids model introduced in 1986 by Reynolds. In that case the SPP are point particles, which move with a constant speed. and adopt (at each time increment) the average direction of motion of the other particles in their local neighborhood up to some added noise.
Simulations demonstrate that a suitable "nearest neighbour rule" eventually results in all the particles swarming together, or moving in the same direction. This emerges, even though there is no centralised coordination, and even though the neighbours for each particle constantly change over time (see the interactive simulation in the box on the right).
Since then a number of models have been proposed, ranging from the simples so called Active Brownian Particle to highly elaborated and specialized models aiming at describing specific systems and situations. Among the important ingredients in these models, one can list
One can also include effective influences of the surrounding; for instance the nominal velocity of the SPP can be set to depend on the local density, in order to take into account crowding effects.
Marching locusts
Young desert locusts are solitary and wingless nymphs. If food is short they can gather together and start occupying neighbouring areas, recruiting more locusts. Eventually they can become a marching army extending over many kilometres. This can be the prelude to the development of the vast flying adult locust swarms which devastate vegetation on a continental scale.
One of the key predictions of the SPP model is that as the population density of a group increases, an abrupt transition occurs from individuals moving in relatively disordered and independent ways within the group to the group moving as a highly aligned whole. Thus, in the case of young desert locusts, a trigger point should occur which turns disorganised and dispersed locusts into a coordinated marching army. When the critical population density is reached, the insects should start marching together in a stable way and in the same direction.
In 2006, a group of researchers examined how this model held up in the laboratory. Locusts were placed in a circular arena, and their movements were tracked with computer software. At low densities, below 18 locusts per square metre, the locusts mill about in a disordered way. At intermediate densities, they start falling into line and marching together, punctuated by abrupt but coordinated changes in direction. However, when densities reached a critical value at about 74 locusts/m2, the locusts ceased making rapid and spontaneous changes in direction, and instead marched steadily in the same direction for the full eight hours of the experiment (see video on the left). This confirmed the behaviour predicted by the SPP models.
In the field, according to the Food and Agriculture Organization of the United Nations, the average density of marching bands is 50 locusts/m2 (50 million locusts/km2), with a typical range from 20 to 120 locusts/m2. The research findings discussed above demonstrate the dynamic instability that is present at the lower locust densities typical in the field, where marching groups randomly switch direction without any external perturbation. Understanding this phenomenon, together with the switch to fully coordinated marching at higher densities, is essential if the swarming of desert locusts is to be controlled.
Bird landings
Swarming animals, such as ants, bees, fish and birds, are often observed suddenly switching from one state to another. For example, birds abruptly switch from a flying state to a landing state. Or fish switch from schooling in one direction to schooling in another direction. Such state switches can occur with astonishing speed and synchronicity, as though all the members in the group made a unanimous decision at the same moment. Phenomena like these have long puzzled researchers.
In 2010, Bhattacharya and Vicsek used an SPP model to analyse what is happening here. As a paradigm, they considered how flying birds arrive at a collective decision to make a sudden and synchronised change to land. The birds, such as the starlings in the image on the right, have no decision-making leader, yet the flock know exactly how to land in a unified way. The need for the group to land overrides deviating intentions by individual birds. The particle model found that the collective shift to landing depends on perturbations that apply to the individual birds, such as where the birds are in the flock. It is behaviour that can be compared with the way that sand avalanches, if it is piled up, before the point at which symmetric and carefully placed grains would avalanche, because the fluctuations become increasingly non-linear.
"Our main motivation was to better understand something which is puzzling and out there in nature, especially in cases involving the stopping or starting of a collective behavioural pattern in a group of people or animals ... We propose a simple model for a system whose members have the tendency to follow the others both in space and in their state of mind concerning a decision about stopping an activity. This is a very general model, which can be applied to similar situations." The model could also be applied to a swarm of unmanned drones, to initiating a desired motion in a crowd of people, or to interpreting group patterns when stock market shares are bought or sold.
Other examples
SPP models have been applied in many other areas, such as schooling fish, robotic swarms, molecular motors, the development of human stampedes and the evolution of human trails in urban green spaces. SPP in Stokes flow, such as Janus particles, are often modeled by the squirmer model,.