Swarm behavior, or swarming, is when many similar-sized animals or objects come together and move in groups. They may stay in one place, move together in large numbers, or travel in a particular direction. This topic involves many different areas of study, such as biology, physics, and computer science.

The word "swarm" is often used for insects, but it can also describe other animals or even non-living things. For example, "flocking" or "murmuration" refers to birds moving together, "herding" describes group movement in four-legged animals, and "shoaling" or "schooling" refers to fish swimming together. Phytoplankton, which are tiny algae, sometimes form large groups called blooms, even though they do not move on their own like most animals. The term "swarm" is also used for non-living things, such as robot swarms, earthquake swarms, or star swarms.

From a scientific perspective, swarm behavior involves many self-moving objects moving together. Mathematicians study this as a result of simple rules followed by each individual, without any leader directing the group. Physicists who study active matter also examine swarming as a phenomenon that does not follow the usual rules of thermodynamic equilibrium. This has led scientists to compare swarming to the behavior of superfluids, such as the way starling flocks move together.

Swarm behavior was first shown in a computer simulation in 1986 using a program called "boids." This program models simple moving agents (called boids) that follow basic rules to create group movement. The simulation was originally created to show how birds flock, but it can also be used to study how fish school or other swarming behaviors.

Models

Scientists have studied how groups of animals move together, called swarm behavior, to learn more about their actions. Early research used math to create models that showed how animals in a group move. These models usually follow three simple rules:
– Move in the same direction as nearby animals.
– Stay close to nearby animals.
– Avoid bumping into nearby animals.

A computer program called "boids," made by Craig Reynolds in 1986, showed how animals move by following these rules. Later models often use "zones" around each animal to guide their movement. In the "zone of repulsion," very close to an animal, it moves away to avoid collisions. In the "zone of alignment," slightly farther away, it matches the direction of nearby animals. In the "zone of attraction," the farthest zone, it moves toward nearby animals.

The shape of these zones depends on how animals sense their surroundings. For example, birds cannot see behind their bodies, while fish use both sight and special body parts called lateral lines. Antarctic krill use sight and antennae to sense movement in water.

Recent studies of starling flocks show that each bird adjusts its position based on the six or seven birds around it, no matter how close or far they are. This suggests that starlings follow a "topological" rule, not a "metric" one. It is still unknown if other animals use this rule. Another study used high-speed camera footage of flocks in Rome to simulate flock behavior using simple rules.

To understand why animals form swarms, scientists use models that simulate how animals evolve. These models often use a method called a genetic algorithm to test how swarming behaviors might develop. Studies have looked at theories like the "selfish herd" idea, the "predator confusion" effect, and the "dilution effect" to explain why swarming helps animals survive.

A concept called "emergence" explains how complex behaviors appear in systems made up of simple parts. For example, ant colonies show complex behaviors even though the queen does not give direct orders. Each ant reacts to chemical signals from other ants, food, or waste, leaving trails that guide others. These trails create patterns that lead to organized group actions, like solving geometric problems.

Another key idea in swarm intelligence is "stigmergy," which means actions leave traces that influence future actions. For example, ants leave pheromone trails that guide other ants to food. This process allows simple actions to build complex, organized results without direct communication.

Swarm intelligence describes how groups of simple agents, like ants or robots, work together without a leader. The term was first used in 1989 to describe robotic systems. These systems are studied in both natural and artificial settings. Research in this field includes scientists studying how swarms work and engineers using these ideas to solve real-world problems.

Swarm models can use two main approaches: the "Lagrangian" method, which tracks individual agents, or the "Eulerian" method, which looks at the whole group as a field. The Lagrangian method is more common because it can track details like direction and spacing that the Eulerian method misses.

Ant colony optimization is a problem-solving method inspired by ants. It uses the way ants leave pheromone trails to find the shortest path to food. This method was created in 1992 and has been used to solve many types of problems. Some ant species form new colonies when a queen and workers leave the original nest, similar to how bees swarm.

Ants are simple individually but can do complex tasks together. They use chemical signals called pheromones to communicate. For example, ants lay trails that help others find food, and they use different pheromones to determine the shortest path between places.

The idea of "self-propelled particles" was introduced in 1995. This model is similar to the "boids" model but focuses on particles that move at a constant speed and adjust their direction based on the movement of nearby particles. Simulations show how these particles can form organized group behaviors.

Biological swarming

The earliest signs of swarm behavior in animals are about 480 million years old. Fossils of a trilobite called Ampyx priscus show that these ancient creatures were often found in lines along the ocean floor. All the trilobites were fully grown adults, and they all faced the same direction, like a line of people walking together. Scientists think they may have lined up to move together, similar to how spiny lobsters travel in single-file groups. Others believe the line might have helped them find mates, like how a type of fly called Leptoconops torrens behaves. These discoveries show that animals have been using group behaviors for a very long time in their evolution.

Swarm behavior is seen in many living things, including bird flocks, fish schools, insect swarms, bacteria, molds, molecular motors, groups of four-legged animals, and even people.

Social insects, such as ants, bees, wasps, and termites, live in large groups and have always interested people, from children to scientists and artists. Each insect seems to act on its own, but the whole group works together in an organized way. Scientists have found that this teamwork happens naturally, without a leader. The way the group acts is often the result of simple interactions between individuals. For example, one ant might follow a trail left by another. These small actions, when added together, help the group solve complex problems, like finding the shortest path to food. This kind of group behavior is sometimes called "swarm intelligence," a type of natural pattern that emerges from simple rules.

Individual ants do not act in complicated ways, but together, they can do things like build nests, care for their young, create bridges, and search for food. A group of ants can choose the best food source by using simple rules. When an ant finds food, it returns to the nest and leaves a chemical trail. The stronger the food, the more chemical is left behind. If two food sources are equally far away but one is better, the stronger trail leads to the better food. Other ants follow the stronger trail, so more ants go to the better food. This creates a cycle where more ants go to the best food, helping the group make a decision. Ants also choose the shortest path to food because the ants that return first are more likely to have taken the shorter route. More ants then follow that path, reinforcing the trail.

Army ants are different from most ants because they do not build permanent nests. Instead, they move constantly, staying in a state of swarming almost all the time. Army ants have evolved similar behaviors in different species, which scientists call "legionary behavior." This may be an example of "convergent evolution," where similar traits develop independently in different groups.

Scientists have studied how ant colonies work to create systems in computer science and robotics that solve problems. These systems are designed to be flexible and able to handle mistakes. This area of study, called "biomimetics," has led to research on how ants move, how search engines use "foraging trails," and how networks can be made more reliable.

In temperate climates, honey bees form swarms in late spring. A swarm usually includes about half the worker bees and the old queen, while the new queen stays behind with the rest of the workers. When bees leave the hive to form a swarm, they often gather on a tree branch or bush near the hive. The bees cluster around the queen, and about 20 to 50 scouts search for a new home. These scouts are the most experienced foragers. If a scout finds a good location, it returns to the group and performs a dance called the "waggle dance." This dance tells other bees about the quality, direction, and distance of the new site. The more excited the scout is, the more vigorously it dances. If other bees agree, they may check the site and promote it. Scouts often compare several sites, sometimes choosing a better site than their own. After some time, the group agrees on the best location. If no decision is made, the swarm may split, and both groups may fail. A good new home needs to be large enough for the swarm (about 15 liters), well-protected, receive enough sunlight, be high off the ground, have a small entrance, and resist ants. That is why tree hollows are often chosen.

Unlike social insects, swarms of non-social insects are usually seen in situations like mating, feeding, avoiding predators, or migrating.

Moths sometimes mate in groups. Female moths release chemicals called pheromones to attract males. Males use their sensitive antennae to detect these chemicals and may follow them for several kilometers. In this type of mating, the first male to reach the female usually mates with her. Females control when and how much pheromone they release to balance attracting mates and avoiding less fit males. After mating, females lay eggs on a host plant. The quality of the host plant can influence where the moths mate and lay eggs. For example, researchers saw pink-striped oakworm moths gather near a dead animal, where the soil might be richer and the plants better.

Midges, like Tokunagayusurika akamusi, form swarms in the air, often dancing together. These swarms help males attract females, a behavior called "lek mating." Swarms usually form in the early evening near bushes, hills, water, or even people. This behavior is not instinctive

People

A group of people can sometimes move together in a way similar to swarms, such as when crowds walk or soldiers move together. In Cologne, Germany, two scientists from the University of Leeds showed that people can act like a flock of birds. In their study, if five percent of a group changed direction, the rest of the group followed. When one person was shown as a "predator" and others avoided them, the group moved like a school of fish. Understanding how people move in crowds is important to help prevent injuries at places like football stadiums, concerts, and subway stations.

Mathematical models of flocking behavior are often used in movies to create realistic crowd movements. These models have been used in films to show large groups of people or animals moving naturally. For example, the movie Batman Returns was the first to use swarm technology to show bats moving together. The Lord of the Rings films used a similar system called "Massive" to create realistic battle scenes. Swarm technology is useful because it is inexpensive, reliable, and easy to use.

Computer simulations based on ant behavior have also been used to study how people board airplanes. Airlines use similar ideas to assign planes to airport gates. A system created by Douglas A. Lawson uses swarm theory, which is the idea that a group of ants works better together than one ant alone. In this system, each pilot acts like an ant searching for the best gate. Pilots learn from their experiences, and their choices help the airline operate efficiently. The program can even warn pilots about possible delays before they happen.

Swarm behavior is also seen in traffic patterns, such as when cars move in waves. Scientists have studied how ants move in both directions on trails to better understand how people and vehicles move in crowds. Simulations based on these studies have been used to model situations like stampedes caused by panic.

In marketing, "herd behavior" describes how people are influenced by others when making purchases. A recent conference in Rome discussed how businesses can use this idea to encourage more sales. For example, people may buy more of a product if they see others buying it. Technologies like smart cards and radio frequency identification tags help share information about product popularity. A model called "swarm-moves" was developed to help stores increase sales without offering discounts.

• Helbing D, Keltsch J, Molnar P (1997). "Modelling the evolution of human trail systems". Nature. 388 (6637): 47–50. arXiv: cond-mat/9805158. Bibcode: 1997Natur.388…47H. doi: 10.1038/40353. PMID 9214501. S2CID 4364517.
• Helbing D, Farkas I, Vicsek T (2000). "Simulating dynamical features of escape panic". Nature. 407 (6803): 487–490. arXiv: cond-mat/0009448. Bibcode: 2000Natur.407..487H. doi: 10.1038/35035023. PMID 11028994. S2CID 310346.
• Helbing D, Farkas IJ, Vicsek T (2000). "Freezing by heating in a driven mesoscopic system". Physical Review Letters. 84 (6): 1240–1243. arXiv: cond-mat/9904326. Bibcode: 2000PhRvL..84.1240H. doi: 10.1103/PhysRevLett.84.1240. PMID 11017488. S2CID 18649078.

Robotics

Swarm robotics is the use of robot groups inspired by the behavior of insect colonies, such as ants and bees. These groups consist of many small robots that work together to complete tasks, like searching for objects, cleaning, or gathering information. Each robot is simple, but when they act together, they create complex group behavior. A group of robots can be seen as a single system, similar to how an ant colony functions as a superorganism that shows swarm intelligence. The largest robot swarm created so far includes 1,024 robots called Kilobots. Other large swarms include the iRobot swarm, the SRI International/ActivMedia Robotics Centibots project, and the Open-source Micro-robotic Project swarm. These swarms are studied to understand how groups of robots can work together. Swarms are also more reliable because if some robots fail, the rest can still complete the task. This makes them useful for space missions, where failures are costly. Swarm robotics also includes research on flying robots and teams that mix ground and flying robots.

At the microscopic level, tiny particles called colloidal particles can act as agents to perform tasks through mechanical and physical methods. These particles can form groups that mimic behaviors seen in nature, such as fish schooling or predator behavior in mammals. They can also move objects by changing shape or organizing themselves. These particles are usually created using chemical processes.

Military

Military swarming is a tactic where groups of units, such as soldiers or machines, attack an enemy from many different directions and then come back together. Pulsing, which means changing where the attack happens, is also part of this tactic. Swarming uses groups that work independently but still communicate and coordinate with each other. These groups focus on moving quickly, staying in touch, working on their own, and working together. In the past, armies used these tactics without studying them closely, but now researchers are studying how swarming ideas can be used in military plans.

Just because many units attack the same target, it does not always mean they are swarming. Siege operations are not swarming because the units do not move around; they all attack the same place, like a fortress. Guerrilla ambushes are not swarms either because they attack quickly and then leave, even if they attack from different places.

In 2014, the U.S. Office of Naval Research showed a video of tests with small boats that can move on their own and work together to attack as a group.

Gallery

• A group of moving herrings
• A group of bees
• Salps that form chains create large groups.
• People moving quickly through an exit do not always act like a liquid.
• A group of ladybirds
• A group of robots
• A group of earthquakes
• A group of ancient stars

Myths

• A common belief is that lemmings die in large groups by jumping off cliffs during their migration. In reality, some lemming species move in large groups when their population becomes too large. Lemmings can swim and may attempt to cross water to find a new home. If the water is too wide for them to cross safely, many may drown. This behavior, along with changes in the numbers of Norwegian lemmings, led to the creation of the myth.
• Piranhas are often thought of as fearless fish that hunt together in groups. However, recent studies, which began by assuming piranhas school to hunt together, found that they actually form groups to protect themselves from predators like cormorants, caimans, and dolphins. A researcher described piranhas as "similar to regular fish, but with large teeth."