Chapter 1: Swarm intelligence

The collective behavior of decentralized, self-organized systems, either natural or artificial, is what is referred to as swarm intelligence (SI). The idea is used in research that is being done on artificial intelligence. In 1989, Gerardo Beni and Jing Wang were the ones who first used the phrase "cellular robotic systems" in connection with their respective fields of study. In natural systems, swarm intelligence may be seen in things like ant colonies, bee colonies, bird flocking, hawk hunting, animal herding, bacterial growth, fish schooling, and microbial intelligence. Other examples include animal herding.

Swarm robotics refers to the application of swarm principles to robots, while swarm intelligence refers to the more broad collection of algorithms that may be used in swarm robotics. The concept of swarm prediction has been used to the resolution of forecasting issues. In the field of synthetic collective intelligence, similar techniques to those that have been suggested for swarm robots are being studied for genetically engineered animals.

Boids is a software that replicates the behavior of birds flocking together that was created by Craig Reynolds in 1986 as part of an artificial life system. His article on this subject was included in the proceedings of the ACM SIGGRAPH conference in 1987 and was published there. A bird-like thing is referred to as a "boid," which is an abbreviated form of the phrase "bird-oid object." The word "boid" relates to this reduced version.

Boids is an example of emergent behavior, which is to say that the complexity of Boids comes from the interaction of individual agents (the boids, in this instance), each of which adheres to a set of basic rules. This is true of the majority of artificial life simulations. The following guidelines govern behavior in the most basic form of the Boids world::

separating: steer to avoid squeezing in with the other nearby sheep.

alignment: go in the general direction that the majority of nearby flockmates are going.

cohesion: migrate toward the average location (center of mass) of nearby flockmates by following the leader's instructions to do so.

It's possible to throw in more complicated rules, such avoiding obstacles and working toward your goals.

Self-propelled particles, or SPP for short, were first described by Vicsek and colleagues in 1995. This model is also known as the Vicsek model.

The majority of work in the subject of nature-inspired metaheuristics is done using evolutionary algorithms (EA), particle swarm optimization (PSO), differential evolution (DE), and ant colony optimization (ACO), as well as its derivatives. The algorithms released up to about the year 2000 are included in this list. The research community has begun to criticize a significant number of more recent metaphor-inspired metaheuristics for masking their lack of originality behind an intricate metaphor. This criticism is directed at a big number of metaphor-inspired metaheuristics. See the list of metaphor-based metaheuristics for any algorithms that have been published after that period.

The use of metaheuristics is not recommended because of its lack of reliability.

The first edition was released in 1989. Search based on stochastic diffusion (SDS)

Ant colony optimization (ACO) is a family of optimization algorithms that is patterned on the activities of an ant colony. Ant colony optimization was first developed by Dorigo in his PhD dissertation. The ACO algorithm is a probabilistic method that may be helpful in solving issues that involve finding better pathways across graphs. Artificial 'ants,' also known as simulation agents, are used to identify optimum solutions by traversing a parameter space that represents all of the potential solutions. During the process of exploring their habitat, natural ants leave behind pheromone trails that guide other ants to resources. The simulated 'ants' also keep track of their locations and the quality of the solutions they come up with, which allows for improved results in following rounds of the simulation by having more ants look for better options.

Particle swarm optimization, often known as PSO, is a method for dealing with issues in which the optimal solution may be represented as a point or surface in an n-dimensional space. PSO is considered a global optimization technique. The hypotheses are mapped out in this space and seeded with an initial velocity. Additionally, a communication channel is established between the different particles. After that, particles will travel across the solution space, and at the end of each timestep, they will be assessed in accordance with some fitness criteria. Particles, over the course of time, are gradually propelled towards the direction of other particles within their communication grouping that have higher fitness values. The large number of members that comprise the particle swarm render the technique impressively resistant to the issue of local minima, which is the primary benefit of such an approach in comparison to other global minimization strategies such as simulated annealing. Other global minimization strategies include:.

The term "Artificial Swarm Intelligence," or ASI, refers to a way of increasing the collective intelligence of networked human groups by using control algorithms that are patterned after the behavior of natural swarms. The technology, which unites groups of human participants into real-time systems that discuss and settle on answers as dynamic swarms when simultaneously presented with a query, is often referred to as Human Swarming or Swarm AI.

Techniques based on swarm intelligence may be put to use in a variety of different applications. For the purpose of managing autonomous vehicles, the United States military is looking at swarm tactics. The European Space Agency is contemplating the use of an orbiting swarm for the purposes of interferometry and self-assembly. The employment of swarm technology for the purpose of planetary mapping is something that NASA is looking at. M. Anthony Lewis and George A. Bekey published a study in 1992 in which they discussed the prospect of using swarm intelligence to manage nanobots inside the body for the aim of destroying cancer tumors.

Research has also been done into the use of swarm intelligence in the field of communications networks, namely in the form of ant-based routing. During the middle of the 1990s, this was first developed independently by Dorigo et al. and by Hewlett Packard. Since then, a variety of other variations have been developed. In essence, this makes use of a probabilistic routing table that rewards and reinforces the path that has been successfully travelled by each "ant," which is a tiny control packet that floods the network. Research has been conducted on the route's reinforcement in both the forwards and backwards directions, as well as in both directions at the same time. Backwards reinforcement requires a symmetric network and couples the two directions together, whereas forwards reinforcement rewards a route before the outcome is known (but then one would pay for the cinema before one knows how good the film is). As a result of the system's stochastic behavior and, as a consequence, its lack of repeatability, there are significant barriers to the commercial adoption of the technology. Because of swarm intelligence, mobile media and other newly developed technologies have the ability to alter the threshold for collective action (Rheingold: 2002, P175).

The optimal placement of the transmission equipment for wireless communication networks is a significant technical challenge that involves balancing a number of conflicting goals. Under the condition that appropriate area coverage is provided for users, just a limited selection of locations (or sites) is necessary. Stochastic diffusion search (SDS), a swarm intelligence technique that was quite differently inspired by ants, has been utilized effectively to give a generic model for this issue, which is linked to circle packing and set covering. It has been shown that the SDS may be used to locate appropriate answers notwithstanding the scale of the issue being investigated.

Swarm technology is being used by artists as a method for the creation of complicated interactive systems or the simulation of crowds.

During the fight sequences of the Lord of the Rings film trilogy, a comparable piece of technology called Massive (software) was used. Swarm technology is especially appealing due to the fact that it is not only inexpensive but also easy and reliable.

Breaking the Ice was the first film to use swarm technology for rendering, and it was the first film to accurately represent the motions of groups of fish and birds using the Boids system. Stanley and Stella starred in the film.

In the film Batman Returns, directed by Tim Burton, swarm technology was again used to demonstrate the flight patterns of a colony of bats.

In order to model the process of people boarding an aircraft, airlines have turned to swarm theory. Researcher Douglas A. Lawson of Southwest Airlines used an ant-based computer simulation using just six interaction rules in order to analyze boarding times utilizing a variety of boarding approaches. (Miller, 2010, xii-xviii).

Through the use of real-time closed-loop control systems, networks of dispersed users have the potential to be organized into "human swarms.".

Stochastic grammars are the building blocks of swarm grammars, which are swarms of stochastic grammars that may be developed to represent complex features such as those found in art and architecture.

In a series of papers, al-Rifaie et al. proposes a revolutionary strategy that deploys the mechanism of 'attention' by...