Swarm Intelligence


Swarm intelligence (SI) involves multiple simple agents interacting with each other and the environment to solve complex problems through their collective global behavior. This is

inspired by the intelligent behavior seen in swarms of animals such as a colony of ants, flocks of birds or schools of fish. It is an artificial intelligence technique based around the study of collective behavior in decentralized, self-organized, systems. Such systems are typically made up of a population of simple agents interacting locally with one another and with their environment. Although there is normally no centralized control structure dictating how individual agents should behave, local interactions between such agents often lead to the emergence of global behavior. It is the discipline that deals with natural and artificial systems composed of many individuals that coordinate using decentralized control and self-organization. In particular, the discipline focuses on the collective behaviors that result from the local interactions of the individuals with each other and with their environment. Examples of systems studied by swarm intelligence are colonies of ants and termites, schools of fish, flocks of birds, herds of land animals. Some human artifacts also fall into the domain of swarm intelligence, notably some multi-robot systems, and also certain computer programs that are written to tackle optimization and data analysis problems.

GIT Jaipur


Swarm Intelligence

Swarm intelligence has a marked multidisciplinary character since systems with the above mentioned characteristics can be observed in a variety of domains. Research in swarm intelligence can be classified according to different criteria.

1. Natural vs. Artificial: It is customary to divide swarm intelligence research into two areas according to the nature of the systems under analysis. We speak therefore of natural swarm intelligence research, where biological systems are studied; and of artificial swarm intelligence, where human artifacts are studied.

2. Scientific vs. Engineering: An alternative and somehow more informative classification of swarm intelligence research can be given based on the goals that are pursued: we can identify a scientific and an engineering stream. The goal of the scientific stream is to model swarm intelligence systems and to single out and understand the mechanisms that allow a system as a whole to behave in a coordinated way as a result of local individualindividual and individual-environment interactions. On the other hand, the goal of the engineering stream is to exploit the understanding developed by the scientific stream in order to design systems that are able to solve problems of practical relevance.

The two dichotomies natural/artificial and scientific/engineering are orthogonal: although the typical scientific investigation concerns natural systems and the typical engineering application concerns the development of an artificial system, a number of swarm intelligence studies have been performed with swarms of robots for validating mathematical models of biological systems. These studies are of a merely speculative nature and definitely belong in the scientific stream of swarm intelligence. On the other hand, one could influence or modify the behavior of the individuals in a biological swarm so that a new swarm-level behavior emerges that is functional to the solution of some task of practical interest. In this case, although the system at hand is a

GIT Jaipur


Deneubourg and his group showed that. Some scientists have programmed a group of robots to implement a similar clustering behavior demonstrating in this way one of the first swarm intelligence scientific oriented studies in which artificial agents were used. y Artificial/Scientific: Clustering by a Swarm of Robots Several ant species cluster corpses to form cemeteries. In the following. No example is available in this area of swarm intelligence although some promising research is currently in progress: For example. an example is given for each of the four possible cases: y Natural/Scientific: Foraging Behavior of Ants In a now classic experiment. The technology developed within this project could be applied to various domains including agriculture and cattle breeding. when given the choice between two paths of different length joining the nest to a food source. in the Leurre project. the pheromone being deposited by the ants while moving from the nest to the food source and back. small insect-like robots are used as lures to influence the behavior of a group of cockroaches. the goals pursued are definitely those of an engineering application. y Natural/Engineering: Exploitation of collective behaviors of animal societies A possible development of swarm intelligence is the controlled exploitation of the collective behavior of animal societies. GIT Jaipur 3 . ants pick up and drop items with probabilities that depend on information on corpse density which is locally available to the ants. In some models. a colony of ants has a high probability to collectively choose the shorter one.Swarm Intelligence natural one. Deneubourg has shown that this behavior can be explained via a simple probabilistic model in which each ant decides where to go by taking random decisions based on the intensity of pheromone perceived on the ground.

A seminal work in this direction was undertaken by Lumer and Faieta in 1994. The same algorithm has also been used for solving combinatorial optimization problems reformulated as clustering problems (Bonabeau et al. 1999).Swarm Intelligence y Artificial/Engineering: Swarm-based Data Analysis Engineers have used the models of the clustering behavior of ants as an inspiration for designing data mining algorithms. They defined an artificial environment in which artificial ants pick up and drop data items with probabilities that are governed by the similarities of other data items already present in their neighborhood. GIT Jaipur 4 .

e. There may be many other properties of a typical swarm intelligence system: y y It is composed of many individuals. The individuals are relatively homogeneous (i. that is. y The overall behavior of the system results from the interactions of individuals with each other and with their environment. These features allow SI systems to handle many problems that are not suitable by traditional mean. the group behavior self-organizes. y Robust: As the solution given is the global response this means the system will still give the correct response even if individual agents fail. non predictable. they are either all identical or they belong to a few typologies).. not defined or computationally hard.Swarm Intelligence CHAPTER-3 FEATURES OF SWARM INTELLIGENCE SI systems have a number of features: y Flexibility: The system is constantly updating means it can respond to changes in the environment. y The interactions among the individuals are based on simple behavioral rules that exploit only local information that the individuals exchange directly or via the environment. y Simplicity: Individual agents are typically very simple systems. y Decentralized: There is no hierarchy amongst the agents and central control system. y Self Organizing: Solutions to a system are emergent rather than predefined. GIT Jaipur 5 . These include problems that are dynamic.

the behavior of each individual of the swarm is described in probabilistic terms: Each individual has a stochastic behavior that depends on his local perception of the neighborhood GIT Jaipur 6 . This is the result of the interaction of spatially neighboring individuals that act on the basis of simple rules. Many examples can be observed in nature of swarms that perform some collective behavior without any individual controlling the group. or being aware of the overall group behavior. the swarm as a whole can show an intelligent behavior.Swarm Intelligence The characterizing property of a swarm intelligence system is its ability to act in a coordinated way without the presence of a coordinator or of an external controller. Notwithstanding the lack of individuals in charge of the group. Most often.

the complexity of Boids arises from the interaction of individual agents (the boids. His paper on this topic was published in 1987 in the proceedings of the ACM SIGGRAPH conference. Boids is an example of emergent behavior. developed by Craig Reyolds in 1986.1. The name refers to a "bird-like object".1 Boids: Boids is an artificial life program.1 Separation: Steer to avoid crowding local flock mates Separation GIT Jaipur 7 . but its pronunciation evokes that of "bird" in a stereotypical New York accent.Swarm Intelligence CHAPTER-4 SOME EXAMPLES AND ALGORITHMS OF SWARM INTELLIGENCE 4. As with most artificial life simulations. which simulates the flocking behaviour of birds. that is. The rules applied in the simplest Boids world are as follows: 4. in this case) adhering to a set of simple rules.

The boids framework is often used in computer graphics. providing realistic-looking representations of flocks of birds and other creatures. GIT Jaipur 8 .1. such as schools of fish or herds of animals. The movement of Boids can be characterized as either chaotic (splitting groups and wild behavior) or orderly. It was for instance used in the 1998 video game Half-Life for the flying bird-like creatures seen at the end of the game on Xen. Unexpected behaviors. can be considered emergent. such as obstacle avoidance and goal seeking. such as splitting flocks and reuniting after avoiding obstacles.Swarm Intelligence 4. named "boid" in the game files.3 Cohesion: Steer to move toward the average position (center of mass) of local flock mates Cohesion More complex rules can be added.1.2 Alignment: Steer towards the average heading of local flock mates Alignment 4.

GIT Jaipur 9 . Ants will randomly spread out and search for food.1 Introduction ACO replicate the natural behavior of ants. ACO has been applied successfully to many classical combinatorial optimization problems. as well as to discrete optimization problems that have stochastic and/or dynamic components. The artificial ants incrementally build solutions by moving on the graph. a set of parameters associated with graph components (either nodes or edges) the values of which are modified at runtime by the ants. Introducing new ants randomly over time allows responses to dynamic changes in the environment. The ACO algorithm finds optimal solutions because shorter paths are traveled over faster and hence more often quickly leading to strong pheromone trails.Swarm Intelligence 4. In ant colony optimization (ACO).2 Ant Colony Optimization 4. that is. Maniezzo and Colorni 1991. Dorigo and Stützle 2004) is a population-based metaheuristic that can be used to find approximate solutions to difficult optimization problems. When food is discovered an ant will return to its base leaving a pheromone trail. The solution construction process is stochastic and is biased by a pheromone model.2. Ant colony optimization (Dorigo. ACO is typically used to find an optimal path. a set of software agents called "artificial ants" search for good solutions to a given optimization problem transformed into the problem of finding the minimum cost path on a weighted graph.

The idea of the ant colony algorithm is to mimic this behavior with "simulated ants" walking around the graph representing the problem to solve. If other ants find such a path. when one ant finds a good (i. thus reducing its attractive strength. The more time it takes for an ant to travel down the path and back again. then returns to the nest (N). and thus the pheromone density remains high as it is laid on the path as fast as it can evaporate. they are likely not to keep travelling at random. GIT Jaipur 10 . in which ants¶ individually limited cognitive abilities have collectively been able to find the shortest path between a food source and the nest. gets marched over faster. The original idea comes from observing the exploitation of food resources among ants. A short path. via any way (a). and upon finding food return to their colony while laying down pheromone trails. Thus. If there were no evaporation at all. ants (initially) wander randomly. other ants are more likely to follow that path.  The first ant finds the food source (F).2 Details of ACO In the natural world. but to instead follow the trail. Over time. leaving behind a trail pheromone (b)  Ants indiscriminately follow four possible ways.2. the paths chosen by the first ants would tend to be excessively attractive to the following ones. the exploration of the solution space would be constrained. the pheromone trail starts to evaporate. however. and positive feedback eventually leads all the ants following a single path.. by comparison. Pheromone evaporation has also the advantage of avoiding the convergence to a locally optimal solution.  Ants take the shortest route.e.Swarm Intelligence 4. long portions of other ways lose their trail pheromones. short) path from the colony to a food source. the more time the pheromones have to evaporate. but the strengthening of the runway makes it more attractive as the shortest route. In that case. returning and reinforcing it if they eventually find food.

because of feedback. in a given amount of time. it returns more or less directly to the nest. These pheromones are attractive. 6. 2. leaving in its path a trail of pheromone. The short route will be increasingly enhanced. An ant (called "blitz") runs more or less at random around the colony. more or less directly. Theoretically. The long route will eventually disappear because pheromones are volatile. They exchange information indirectly by depositing pheromones. all the ants have determined and therefore "chosen" the shortest route. if the quantity of pheromone remained the same over time on all edges. If there are two routes to reach the same food source then. the track. Eventually. all detailing the status of their "work". to a stable state where the route is composed of the strongest edges. Returning to the colony. and therefore become more attractive. these ants will strengthen the route. 4. only an ant located where the pheromones were left has a notion of them. 3. a slight variation on an edge will be amplified and thus allow the choice of an edge. However. nearby ants will be inclined to follow. The information exchanged has a local scope. A model explaining this behaviour is as follows: 1. Ants use the environment as a medium of communication. 8. GIT Jaipur 11 . biologists have observed that ants tended to use the shortest route. The mechanism to solve a problem too complex to be addressed by single ants is a good example of a self-organized system. no route would be chosen. 5. This system is called "Stigmergy" and occurs in many social animal societies (it has been studied in the case of the construction of pillars in the nests of termites). The algorithm will move from an unstable state in which no edge is stronger than another. This system is based on positive feedback (the deposit of pheromone attracts other ants that will strengthen it themselves) and negative (dissipation of the route by evaporation prevents the system from thrashing). If it discovers a food source. 7. the shorter one will be traveled by more ants than the long route.Swarm Intelligence In a series of experiments on a colony of ants with a choice between two unequal length paths leading to a source of food.

5. produce near-optimal solutions to the travelling salesman problem. the ant chooses to move from one city to another according to some rules: 1. Trail evaporation reduces all trail values over time thereby avoiding any possibilities of getting stuck in local optima. They have an advantage over simulated annealing and genetic algorithm approaches of similar problems when the graph may change dynamically. multi-targets and parallel implementations. each such ant incrementally constructs a solution to the problem. This pheromone information will direct the search of the future ants. It has also been used to produce near-optimal solutions to the travelling salesman problem. It must visit each city exactly once. trails of pheromones evaporate. the ant colony algorithm can be run continuously and adapt to changes in real time. viz. each making one of the possible round-trips along the cities. A distant city has less chance of being chosen (the visibility). Furthermore. Having completed its journey. in which the goal is to find the shortest round-trip to link a series of cities.Swarm Intelligence The basic philosophy of the algorithm involves the movement of a colony of ants through the different states of the problem influenced by two local decision policies.. trail evaporation and daemon actions. viz. if the journey is short.2. This is of interest in network routing and urban transportation systems. The first ACO algorithm was called the Ant system and it was aimed to solve the travelling salesman problem. At each stage. 2. 3. The more intense the pheromone trail laid out on an edge between two cities. the algorithm also includes two more mechanisms. trails and attractiveness. GIT Jaipur 12 . the greater the probability that that edge will be chosen. stochastic problems..3 Applications Ant colony optimization algorithms have been applied to many combinatorial optimization problems and a lot of derived methods have been adapted to dynamic problems in real variables. or during the construction phase. 4. The daemon actions are used to bias the search process from a non-local perspective. the ant evaluates the solution and modifies the trail value on the components used in its solution. After each iteration. The general algorithm is relatively simple and based on a set of ants. 4. When an ant completes a solution. Thereby. the ant deposits more pheromones on all edges it traversed.

-)  But are traced with absurd geometries  You will definitely start ³cutting the corners´  This will make the grass signed GIT Jaipur 13 .4 Positive and Negative Feedbacks Positive Feedback  Pheromones attract more ants  Get them to food  And make them leave pheromones at their turn  This enables the creation of stable path Negative feedback  Pheromone evaporation ensures that useless path will disappear  Also ensure that all possible solutions.3 Swarming In Humans Imagine a park with stone footpaths  That should discourage you to walk on grass .2. are explored in any case 4.Swarm Intelligence 4.

However. Such methods are commonly known as metaheuristics as they make few or no assumptions about the problem being optimized and can search very large spaces of candidate solutions. GIT Jaipur 14 . change over time. PSO can therefore also be used on optimization problems that are partially irregular. etc. noisy. PSO does not use the gradient of the problem being optimized. More specifically. and it will be the shortest path  If a path go to disuse grass will grow again (as in pheromone evaporation) 4. which means PSO does not require for the optimization problem to be differentiable as is required by classic optimization methods such as gradient descent and quasi-newton methods.Swarm Intelligence  Inviting other people to walk in  Until a clear spontaneous footpath emerge. metaheuristics such as PSO do not guarantee an optimal solution is ever found.4 Particle Swarm Intelligence Particle Swarm Optimization (PSO) is a computational method that optimizes a problem by iteratively trying to improve a candidate solution with regard to a given measure of quality.

that a satisfactory solution will eventually be discovered GIT Jaipur 15 . but not guaranteed. A basic variant of the PSO algorithm works by having a population (called a swarm) of candidate solutions (called particles). and moving these particles around in the search-space according to simple mathematical formulae. here dubbed particles. The movements of the particles are guided by their own best known position in the search-space as well as the entire swarm's best known position. The movements of the particles are guided by the best found positions in the searchspace which are updated as better positions are found by the particles. The process is repeated and by doing so it is hoped.Swarm Intelligence PSO optimizes a problem by having a population of candidate solutions. When improved positions are being discovered these will then come to guide the movements of the swarm. These particles are moved around in the search-space according to a few simple formulae.

Bekey discusses the possibility of using swarm intelligence to control nano bots within the body for the purpose of killing cancer tumors. 5. or to simulate crowds in movie special effects. A 1992 paper by M. The European Space Agency is thinking about an orbital swarm for self assembly and interferometry. Stanley and Stella in: Breaking the Ice was the first movie to make use of GIT Jaipur 16 . Anthony Lewis and George A. E. y Complex optimization: problems like scheduling. y Path optimization: Used to solve the typical µtraveling salesman¶ problem. y Crowd control: Used to control the distribution and behaviour of multiple agents over an area. NASA is investigating the use of swarm technology for planetary mapping. collective robotics. Used to solve dynamic.Swarm Intelligence CHAPTER-5 APPLICATIONS OF SWARM INTELLIGENCE Swarm Intelligence is utilized in the following areas: y Crowd simulation: Used to predict crowd movement in transportation problems. nonlinear complex optimization y y Data Analysis Problem Swarm Robotics Swarm Intelligence-based techniques can be used in a number of applications. The U.S. military is investigating swarm techniques for controlling unmanned vehicles.g.1 Crowd simulation Artists are using swarm technology as a means of creating complex interactive systems or simulating crowds.

As the system behaves stochastically and is therefore lacking repeatability.2 Ant-based routing The use of Swarm Intelligence in Telecommunication Networks has also been researched. 5. forwards reinforcement rewards a route before the outcome is known (but then you pay for the cinema before you know how good the film is). robust. Tim Burton's Batman Returns also made use of swarm technology for showing the movements of a group of bats. This was pioneered separately by Dorigo et al. known as Massive. in the form of Ant Based Routing. during battle scenes. Basically this uses a probabilistic routing table rewarding/reinforcing the route successfully traversed by each "ant" (a small control packet) which flood the network. with a number of variations since. Swarm technology is particularly attractive because it is cheap.Swarm Intelligence swarm technology for rendering. reverse direction and both simultaneously have been researched: backwards reinforcement requires a symmetric network and couples the two directions together. there are GIT Jaipur 17 . Southwest Airlines researcher Douglas A. Reinforcement of the route in the forwards. and simple. realistically depicting the movements of groups of fish and birds using the Boids system. The Lord of the Rings film trilogy made use of similar technology. Airlines have used swarm theory to simulate passengers boarding a plane. Lawson used an ant-based computer simulation employing only six interaction rules to evaluate boarding times using various boarding methods. and Hewlett Packard in the mid-1990s.

The program can even alert a pilot of plane back-ups before they happen. but combined. so the size of the swarm can be increased or decreased as needed. No individual in the group is very intelligent or complex. As a group. so we'll have a gate available.3 Swarm Robotics Swarm robotics is an approach to robotics that emphasizes many simple robots instead of a single complex robot. Swarm robotics has been an experimental field. these simple machines cooperate to perform advanced tasks that otherwise would require a more powerful. Mobile media and new technologies have the potential to change the threshold for collective action due to swarm intelligence (Rheingold: 2002. P175). GIT Jaipur 18 . A traditional robot often needs complex components and significant computer processing power to accomplish its assigned tasks. more expensive robot. they can perform difficult tasks. Robot swarms have high fault tolerance.Swarm Intelligence large hurdles to commercial deployment. the "colony" of pilots always go to gates they can arrive and depart quickly." Dr. Lawson explains. or swarm intelligence -.´ 5. "The pilot learns from his experience what the best is for him. and it turns out that that's the best solution for the airline.Douglas A. but many practical applications have been proposed. Each pilot acts like an ant searching for the best airport gate. each robot is relatively simple and inexpensive. Swarms also are scalable. "We can anticipate that it's going to happen. In swarm robotics. As a result. Using many simple robots has other advantages as well. Airlines have also used ant-based routing in assigning aircraft arrivals to airport gates. meaning that they still will perform well if some of the individual units malfunction or are destroyed. A robot swarm has much in common with an ant colony or swarm of bees.the idea that a colony of ants works better than one alone. At Southwest Airlines software program uses swarm theory.

The data from all of the robots in the swarm is then combined into a single map. GIT Jaipur 19 . swarm robotics emphasizes a large number of robots. like radio or infrared. for instance by using only local communication.g. Other applications include: Optimization problems: ‡ Computer approximate solutions to problems such as combinatorial problems (e.Swarm Intelligence One use that researchers have demonstrated for swarm robotics is mapping. The problems of organizing a swarm haven¶t kept people from imagining what swarm robotics could offer some day. TSP) and routing (e. Some scientists envision a swarm of very small micro bots being used to explore other planets. network routing). A single robot would constantly need to keep track of its location. Unlike distributed robotic systems in general. and promotes scalability. A swarm of robots could be programmed simply to avoid obstacles while keeping in contact with other members of the swarm. mining and even firefighting.g. remember where it had been and figure out how to avoid obstacles while still exploring the entire area. That local communication for example can be achieved by wireless transmission systems. Other proposed uses include search-and-rescue missions.

‡ Self assembling robots.Swarm Intelligence Construction algorithms: ‡ Using self-organization and stigmergy to build structures using a set of simple rules. GIT Jaipur 20 . Cooperative transport: ‡ Accomplishing tasks (in this case transport) that require multiple agents cooperating with each other and without global information. ‡ Robotic cooperative transport.

Some of the most important applications of swarm application framework are: 1. Makes development of swarm applications easier.Swarm Intelligence CHAPTER-6 SWARM INTELLIGENCE FRAMEWORK 6. Provides useful tools and modules. 6. 2.2 Architecture FRAMEWORK LIBRARY USER APPLICATION ENVIRONMENT GIT Jaipur 21 . It provides useful tools and modules that make the development of swarm applications interactive. SI Framework is a type of integrated development environment that makes the development of swarm applications easier and intuitive.1 Introduction Swarm intelligence framework is a software development framework. Makes their development intuitive. 3.

6. 2. In this each agent runs as an independent process.4 Rules These are the parameters that control the agent. tools and everything else. Library: This layer contains useful objects. Framework: This layer lays out the fundamental paradigm for applications in the swarm intelligence framework.3 Agents The compiled result of swarm intelligence framework is called an agent. 6.Swarm Intelligence There are basically three layers of swarm intelligence framework: 1. MAINTAIN EDGE SHAPE FORM CORNER CENTRE ATRRACTION CIRCLE OF RADIUS DISTANCE ATTRACT GIT Jaipur 22 . Application: This layer is where the actual application implementation resides. 3. It is mainly composed of two components:  Rules ±Dictates the behavior of agent by sensing and acting.  Data ±Maintains state that the rules may access.

they tend to be inefficient.  Unpredictable ± The complexity of a swarm system leads to unforeseeable results.Swarm Intelligence CHAPTER-7 LIMITATIONS OF SWARM INTELLIGENCE There are some limitations of swarm intelligence which are mentioned below:  Non-optimal ± Because swarm systems are highly redundant and have no central control. The more complex the swarm. GIT Jaipur 23 . Swarm systems require guidance in the way that a shepherd drives a herd: by applying force at crucial leverage points. Swarms can dampen inefficiency. complex adaptive systems. Instead of A causing B. peripheral players have to come to rest. the longer it takes to shift states. A indirectly causes everything. instead.  Non-immediate ± Linear systems tend to be very direct: Flip a switch and the light comes on. Not all novelty is desirable. but never to the degree that a linear system can. But complex swarm systems with rich hierarchies take time. Simple collective systems tend to operate simply. Emergent novelty is a primary characteristic of self-organization by adaptive systems. and duplication of effort is always rampant. and everything indirectly causes A. and a multitude of autonomous agents need to become acquainted with each other. The allocation of resources is not efficient. Each hierarchical layer has to settle down.  Uncontrollable ± It is very difficult to exercise control over a swarm.  Non-understandable ± Sequential systems are understandable. are a jumble of intersecting logic. which in turn causes C.

bacteria foraging. Natural 2.Swarm Intelligence CONCLUSION Swarm Intelligence (SI) is a computational intelligence technique involving the study of collective behavior in decentralized systems. Such systems are made up by a population of simple individuals interacting locally with one another and with their environment. Particle Swarm Intelligence GIT Jaipur 24 . Although there is typically no centralized control dictating the behavior of the individuals. bird flocking. and many more. Ant Colony Optimisation 3. Engineering Some models and examples of swarm intelligence are: 1. The features of swarm intelligence:  Flexibility  Robust  Decentralized  Self Organizing  Simplicity Swarm intelligence has a marked multidisciplinary character since systems with the above mentioned characteristics can be observed in a variety of domains. local interactions among individuals often cause a global pattern to emerge. animal herding. 1. honey bees. Examples of systems like this can be found in nature. Boids 2. Artificial 4. Scientific 3. fish schooling. Research in swarm intelligence can be classified according to different criteria. including ant colonies.

4116.cs.org/article/Swarm_Intelligence GIT Jaipur 25 .edu/~evans/ion. 2. 3.virginia. 4.org http://arxiv.Swarm Intelligence REFERENCES 1.scholarpedia.pdf http://www.pdf http://www.org/ftp/arxiv/papers/0910/091-0. www.wikipedia.

Sign up to vote on this title
UsefulNot useful