ABSTRACT:

Swarm Intelligence (SI) is a relatively new paradigm being applied in a host of research

settings to improve the management and control of large numbers of interacting entities such

as communication, computer and sensor networks, satellite constellations and more. Attempts

to take advantage of this paradigm and mimic the behavior of insect swarms however often

lead to many different implementations of SI. The rather vague notions of what constitutes self-

organized behavior lead to rather ad hoc approaches that make it difficult to ascertain just what

SI is, assess its true potential and more fully take advantage of it.

This work provides a set of general principles for SI research and development. A precise

definition of self-organized behavior is described and provides the basis for a more axiomatic

and logical approach to research and development as opposed to the more prevalent ad hoc

approach in using SI concepts. The advances and applications of Swarm Intelligence is also dealt

with in this work.

1

CHAPTER ONE

INTRODUCTION1.1 Background of the study

Swarm Intelligence (SI) is the property of a system whereby the collective behaviors of agents

interacting locally with their environment cause coherent functional global patterns to emerge. SI

provides a basis with which it is possible to explore distributed problem solving without

centralized control or the provision of a global model. One of the cores tenets of SI work is that

often a decentralized, bottom-up approach to controlling a system is much more effective than

traditional, centralized approach. Groups performing tasks effectively by using only a small set

of rules for individual behaviour is called swarm intelligence. Swarm Intelligence is a property

of systems of non-intelligent agents exhibiting collectively intelligent behaviour. In Swarm

Intelligence, two individuals interact indirectly when one of them modifies the environment and

the other responds to the new environment at a later time. For years scientists have been studying

about insects like ants, bees, termites etc. The most amazing thing about social insect colonies is

that there’s no individual in charge. For example consider the case of ants. But the way social

insects form highways and other amazing structures such as bridges, chains, nests and can

perform complex tasks is very different: they self-organize through direct and indirect

interactions. The characteristics of social insects are (Bonabeau, 1999.)

1. Flexibility

2. Robustness

3. Self-Organization

2

1.2 Aim and Objectives of the study

The aim of this study is to highlight the most significant recent developments on the topics of

swarm intelligence.

The objectives of this work are

To highlights the area of applications of swarm intelligence

To highlight benefits /advantages of swarm intelligence

To identify future research directions,

To publicize swarm intelligence algorithms to a wider audience.

1.3 Scope of the Study

Swarm intelligence is a relatively new discipline that deals with the study of self-organizing

processes both in nature and in artificial systems. Researchers in ethology and animal behavior

have proposed many models to explain interesting aspects of social insect behavior such as self-

organization and shape-formation. Recently, algorithms inspired by these models have been

proposed to solve difficult computational problems.

An example of a particularly successful research direction in swarm intelligence is ant colony

optimization, the main focus of which is on discrete optimization problems. Ant colony

optimization has been applied successfully to a large number of difficult discrete optimization

problems. Another interesting approach is that of particle swarm optimization, that focuses on

continuous optimization problems. Here too, a number of successful applications can be found in

the recent literature. Swarm robotics is another relevant field. Here, the focus is on applying

swarm intelligence techniques to the control of large groups of cooperating autonomous robots.

3

1.4 Significance of the Study

This study is important to the field of science and engineering as it will improve the applications

the following applications of swarm intelligence;

Particle swarm optimization

Artificial bees and firefly algorithms

Bacterial foraging optimization

Ant colony optimization

Swarm robotics

Artificial immune systems

Hybridization of swarm intelligence methods

Theory and practice of swarm intelligence methods in different domains

Real-world problem solving using swarm intelligence methods

1.5 Definition of terms and abbreviations

Artificial: made or produced by human beings rather than occurring naturally, especially as a

copy of something natural.

Swarm: A large or dense group of flying insects.

Intelligence: The ability to acquire and apply knowledge and skills.

Swarm intelligence: this is the collective behavior of decentralized, self-organized systems,

natural or artificial.

A.C.O: Ant Colony Optimization

CSS: Charged System Search

CHAPTER TWO

4

LITERATURE REVIEW

Swarm intelligence is inspired by emergent behaviour in nature, such as bird flock, ant

colony, and fish school. It is first introduced by Beni and Wang in \Swarm Intelligence in

Cellular Robotic System". In the paper, swarm intelligence is described as \systems of non-

intelligent robots exhibiting collectively intelligent behaviour evident in the ability to

unpredictably produce `specific' ([i.e.] not in a statistical sense) ordered patterns of matter in the

external environment" Beni and Wang (1989).

Craig Reynolds' flocking system is one of the influential swarm intelligence example.

Individuals in flocking system are called boids. They have three layers of motion, action

selection, steering and locomotion. The three simple steering behaviours {aggregation,

separation, alignment {can emergent into complicated result that looks like a flock of bird or a

school of _sh. In his paper, other steering behaviours such as obstacle avoidance, seeking,

foraging are also added to the system to enhance the over all realism. Reynolds (1999)

Figure 2.1: Craig Reynold's flocking system Reynolds (1999)

Another well known example of emergent behaviour would be Conway's Game of Life. It is a

cellular automation with initial state specified and rules are set to decide whether the life live or

die. With different initial state given, different pattern generate after certain time with the three

rules. Conway's idea is a simplified implementation of John Von Neumann's attempt to build a

5

machine that can build itself, which is also an analogy of the rise, fall and alterations of a society

of living organisms Wikipedia (2010).

2.1.1 Properties of Swarm Intelligence

The typical swarm intelligence system has the following properties:

It is composed of many individuals;

The individuals are relatively homogeneous

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

The overall behaviour of the system results from the interactions of individuals with each

other and with their environment, that is, the group behavior self-organizes.

Modelling Swarm Behaviour

The simplest mathematical models of animal swarms generally represent individual

animals as following three rules:

1. Move in the same direction as your neighbor

2. Remain close to your neighbors

3. Avoid collisions with your neighbors

Many current models use variations on these rules, often implementing them by means of

concentric "zones" around each animal. In the zone of repulsion, very close to the animal, the

focal animal will seek to distance itself from its neighbors to avoid collision. Slightly further

away, in the zone of alignment, the focal animal will seek to align its direction of motion with its

neighbors. In the outermost zone of attraction, which extends as far away from the focal animal

as it is able to sense, the focal animal will seek to move towards a neighbor.

6

The shape of these zones will necessarily be affected by the sensory capabilities of the given

animal. For example the visual field of a bird does not extend behind its body. Fish rely on both

vision and on hydrodynamic perceptions relayed through their lateral line, while Antarctic krill

rely both on vision and hydrodynamic signals relayed through antennae. Some of the animals

that exhibit swarm behavior are

1. Insects – Ants, bees, locusts, termites, mosquitoes and insects migration.

2. Bacteria

3. Birds

4. Land animals

5. Aquatic animals – fish, krill and other aquatic animals

6. People

Figure 2.2 : fish ColonySource: Wikipedia.com/swarm-intelligence/fish-colony

2.2 Example Algorithms of Swarm Intelligence:

Ant colony optimization

River formation dynamics

Particle swarm optimization

Stochastic diffusion search

Gravitational search algorithm

7

Intelligent Water Drops

Charged System Search

1. Ant Colony Optimization

In computer science and operations research, the ant colony optimization algorithm

(ACO) is a probabilistic technique for solving computational problems which can be reduced to

finding good paths through graphs. In the real world, ants wander randomly, and upon finding

food return to their colony while laying down pheromone trails. If other ants find such a path,

they are likely not to keep travelling at random, but to instead follow the trail, returning and

reinforcing it if they eventually find food through that way. This algorithm is inspired by

forgiving behavior of the ants.

1. The first ant finds the food source (F), via any way (a), then returns to the nest (N),

leaving behind a trail pheromone (b)

2. Ants indiscriminately follow four possible ways, but the strengthening of the runway

makes it more attractive as the shortest route.

3. Ants take the shortest route, long portions of other ways lose their trail pheromones.

In a series of experiments on a colony of ants with a choice between two unequal length

paths leading to a source of food, biologists have observed that ants tended to use the shortest

route. A model explaining this behavior is as follows:

1. An ant (called "blitz") runs more or less at random around the colony;

2. If it discovers a food source, it returns more or less directly to the nest, leaving in its path

a trail of pheromone;

8

3. These pheromones are attractive, nearby ants will be inclined to follow, more or less

directly, the track;

4. Returning to the colony, these ants will strengthen the route;

5. If there are two routes to reach the same food source then, in a given amount of time, the

shorter one will be traveled by more ants than the long route;

6. The short route will be increasingly enhanced, and therefore become more attractive;

7. The long route will eventually disappear because pheromones are volatile;

8. Eventually, all the ants have determined and therefore "chosen" the shortest route.

Figure 2.3 Ant Colony OptimizationSource: Wikipedia.com/swarm-intelligence/ant-colony

2. River Formation Dynamics

This method is similar to that of the “Ant Colony Optimization”. In fact, this can be seen

as a gradient version of Ant Colony Optimization, based on copying how water forms rivers by

eroding the ground and depositing sediments. The gradients are followed by subsequent drops to

create new gradients, reinforcing the best ones. By doing so, good solutions are given in the form

of decreasing altitudes. This method has been applied to solve different NP-complete problems.

3. Particle Swarm Optimization

9

Particle swarm optimization (PSO) is a global optimization algorithm for dealing with

problems in which a best solution can be represented as a point or surface in an n-dimensional

space. Hypotheses are plotted in this space and seeded with an initial velocity, as well as a

communication channel between the particles. Particles then move through the solution space,

and are evaluated according to some fitness criterion after each time step. Over time, particles

are accelerated towards those particles within their communication grouping which have better

fitness values. The main advantage of such an approach over other global minimization strategies

such as simulated annealing is that the large number of members that make up the particle swarm

make the technique impressively resilient to the problem of local minima.

4. Stochastic Diffusion Search

It belongs to a family of swarm intelligence and naturally inspired search and

optimization algorithms which includes ant colony optimization, particle swarm optimization

and genetic algorithms. It is an agent-based probabilistic global search and optimization

technique best suited to problems where the objective function can be decomposed into multiple

independent partial-functions. Each agent maintains a hypothesis which is iteratively tested by

evaluating a randomly selected partial objective function parameterized by the agent's current

hypothesis.

5. Gravitational Search Algorithm

Gravitational search algorithm (GSA) is constructed based on the law of Gravity and the

notion of mass interactions. The GSA algorithm uses the theory of Newtonian physics and its

searcher agents are the collection of masses. In GSA, we have an isolated system of masses.

10

Using the gravitational force, every mass in the system can see the situation of other masses. The

gravitational force is therefore a way of transferring information between different masses.

6. Intelligent Water Drops

Intelligent Water Drops algorithm (IWD) is a swarm-based nature-inspired optimization

algorithm, which has been inspired from natural rivers and how they find almost optimal paths to

their destination. These near optimal or optimal paths follow from actions and reactions

occurring among the water drops and the water drops with their riverbeds. In the IWD algorithm,

several artificial water drops cooperate to change their environment in such a way that the

optimal path is revealed as the one with the lowest soil on its links. The solutions are

incrementally constructed by the IWD algorithm. Consequently, the IWD algorithm is generally

a constructive population-based optimization algorithm.

7. Charged System Search

Charged System Search (CSS) is a new optimization algorithm based on some principles

from physics and mechanics. CSS is a multi-agent approach in which each agent is a Charged

Particle (CP). CPs can affect each other based on their fitness values and their separation

distances. The quantity of the resultant force is determined by using the electrostatics laws and

the quality of the movement is determined using Newtonian mechanics laws. CSS is applicable

to all optimization fields; especially it is suitable for non-smooth or non-convex domains. This

algorithm provides a good balance between the exploration and the exploitation paradigms of the

11

algorithm which can considerably improve the efficiency of the algorithm and therefore the CSS

also can be considered as a good global and local optimizer simultaneously.

2.3 The wisdom of the anthill

Leading companies are applying the self-organising rules of social insects to make complex

businesses run more efficiently by using so-called agent-based simulations modelled after

individual insects and swarms. The assumption is that if ants, with brains that weigh less than the

ink in this comma, can run efficient supply chains, why do humans have such trouble?

The question has dogged Eric Bonabeau for years. The 34-year-old Frenchman – a scientist and

student of the chaos-theory branch of complexity science – has spent nearly a decade studying

the organisation, co-ordination, and work habits of social insects. Ant colonies are so efficient,

Bonabeau deduced, because they lack centralised control; no single ant boss runs the business.

Bonabeau took this notion a step further in his 1999 book, Swarm Intelligence, in which he

described how the study of an organisation's ‘ants’, its myriad individual parts, could help

businesses find solutions to problems that elude ordinary top-down analysis. (Bonabeau 1999)

For example, how the late arrival of a single package can derail an entire supply chain or why

adding a lane to a highway can often worsen traffic jams.

It all seemed great on paper, but until recently Bonabeau had not had the chance to prove that his

theories would work in practice. When hired by a client company, the consultancies applying

Bonabeau's ideas, notably Santa Fe, New Mexico's BiosGroup, where Bonabeau started, and

Icosystem, the firm he founded after leaving BiosGroup, created algorithms to generate super-

12

realistic simulations of an operation's moving parts. In other words, they plug in virtual

employees, products, and customers; order them to do different tasks; and watch what happens.

Bonabeau's ‘agent-based modelling’ techniques now help to tighten supply chains, speed up the

drugs to market process, and even design better, unmanned drones, for the Pentagon.

Bonabeau's virtual ants, for example, are crawling all over Air Liquide, the French industrial gas

giant. The company supplies liquid oxygen, nitrogen and other gases to some 10,000 customers

from over 300 sources through 30 depots and using 200 trucks and 200 trailers. It is a supply

chain that can create 3trn daily combinations among all its constituent parts. Twenty-two full-

time logistics analysts took nearly half a day to generate a delivery schedule that would move

every product to its destination on time.

Working with BiosGroup, Air Liquide chose to run agent-based simulations to see how they

could draw up more efficient delivery routines. Air Liquide programmed its trucks, like ants, to

find the shortest routes or to follow the equivalent of pheromone trails, which means that

subsequent trucks were to retrace shortcuts that others found. Then, using reams of data from Air

Liquide's business operations, BiosGroup engineers retested their computer simulations until

they found the most efficient combination of rules. The result: just one Air Liquide analyst is

necessary to create daily shipping and production schedules across its numbingly complex

supply chain in about two hours.

Southwest Airlines in the US, too, has benefited from Bonabeau's ideas. The airline has

simulated the various parts of its cargo-shipping business: its aircraft, destinations, types of

cargo, ground personnel, and so on. The simulations showed what Southwest's logistics

managers had suspected all along, it is sometimes more efficient not to take the shortest route, as

13

long as fewer hands actually touch the cargo in the process. Southwest claims that it saves $2m a

year in labour costs because of this insight.

Agent-based modelling has evolved beyond ant behaviour, and so too has Bonabeau's work.

After leaving BiosGroup in 2000, he soon founded Icosystem and began to apply his algorithms

to new business applications. To move Eli Lilly's drugs to market faster, engineers, for example,

built a model of the company's sprawling clinical-development processes, using people, drugs,

regulatory hurdles, and other individual parts to simulate this laborious task. By finding and

applying the right rules, Bonabeau hopes to speed up Eli Lilly's development time by as much as

80 per cent. Insurance firms are also using these methods to better understand how people choose

their own health plans.

Even the US Office of Naval Research is using these techniques to help it design smarter

unmanned aerial vehicles (UAVs). Because a centralised command on the ground controls the

UAVs, they tend not to work well as a team. They bunch up, miss large areas, or fail to respond

to enemy threats. So Bonabeau and his colleagues have created simulations in which virtual

drones follow hundreds of different rules: stay away from other drones, fly to areas no other

drones have covered, and so on. Initially, the goal is to help drones communicate directly with

each other. But by 2020, Pentagon planners hope to create entire swarms of unmanned vehicles

that communicate and attack in concert.

Despite the costs and complexities, more organisations aspire to emulate ant colonies. IBM is

experimenting with its own agent-based modelling programs for its future e-commerce software.

The EU is also funding a three-year modelling research project at the Santa Fe Institute, the

hotbed of complexity science.

14

Beyond modelling, Bonabeau says, the ultimate goal of the technology is to solve problems

before they happen – not unlike, for example, ‘autonomic’ computing systems that seek out and

fix software glitches without any human intervention. But as Bonabeau admits, advancing to the

next phase means that he will have to do better than just deliver more efficient shipping routines

for his customers.

15

CHAPTER THREE

3.1 HOW SWARM INTELLIGENCE IMPACT US

3.1.1 Positive impact on human

It helps in the reduction of man power in companies, firms as swarm robots will be

designed to do the jobs. Example these jobs include assembly of the various parts of auto

mobile

Industrial automated robots have the capacity to dramatically improve product quality.

Applications are performed with precision and high repeatability every time. This level of

consistency can be hard to achieve any other way.

With robots, throughput speeds increase, which directly impacts production. Because an

automated robot has the ability to work at a constant speed without pausing for breaks,

sleep, vacations, it has the potential to produce more than a human worker.

Robots increase workplace safety. Workers are moved to supervisory roles where they no

longer have to perform dangerous applications in hazardous settings.

Improved worker safety leads to financial savings. There are fewer healthcare and

insurance concerns for employers. Automated robots also offer untiring performance

which saves valuable time. Their movements are always exact, minimizing material

waste.

3.1.2 Negative impact

The initial investment to integrated automated robotics into your business is significant,

especially when business owners are limiting their purchases to new robotic equipment.

16

The cost of robotic automation should be calculated in light of a business' greater

financial budget. Regular maintenance needs can have a financial toll as well.

Incorporating industrial robots does not guarantee results. Without planning, companies

can have difficulty achieving their goals.

Employees will require training program and interact with the new robotic equipment.

This normally takes time and financial output.

Robots may protect workers from some hazards, but in the meantime, their very presence

can create other safety problems. These new dangers must be taken into consideration.

3.1.3 Benefits of implementing Swarm Robotics

Robots produce more accurate and high quality work.

Robots rarely make mistakes and are more precise than human workers.

They can produce a greater quantity in a short amount of time.

They can work at a constant speed with no breaks, days off, or holiday time.

Robots save workers from performing dangerous tasks.

They can work in hazardous conditions, such as poor lighting, toxic chemicals, or tight

spaces.

They are capable of lifting heavy loads without injury or tiring.

Robots increase worker safety by preventing accidents since humans are not performing

risky jobs.

Work cells provide safety features, separating the worker from harm’s way.

17

Robots save time by being able to produce a greater magnitude of products

They also reduce the amount of wasted material used due to their accuracy

Robots save companies money in the long run with quick ROIs (return on investment),

fewer worker injuries (reducing or eliminating worker’s comp), and with using less

materials.

3.1.4 CHALLENGES OF IMPLEMENTING SWARM ROBOTICS

(i) Interference: robots in a group can interfere between them, due to collisions,

occlusions, and so forth

(ii) Uncertainty concerning other robots’ intentions: coordination requires to know what

other robots are doing. If this is not clear robots can compete instead of cooperate.

(iii) Overall system cost: the fact of using more than one robot can make the economical

cost bigger. This is ideally not the case of swarm-robotic systems, which intend to use

many cheap and simple robots which total cost is under the cost of a more complex

single robot carrying out the same task.

3.2 OTHER APPLICATION AREA OF SWARM INTELLIGENCE

a) CLUSTERING BEHAVIOR OF ANTS

Ants build cemeteries by collecting dead bodies into a single place in the nest. They also

organize the spatial disposition of larvae into clusters with the younger, smaller larvae in the

cluster center and the older ones at its periphery. This clustering behavior has motivated a

number of scientific studies.

18

b) NEST BUILDING BEHAVIOR OF WASPS AND TERMITES

Wasps build nests with a highly complex internal structure that is well beyond the

cognitive capabilities of a single wasp. Termites build nests whose dimensions are enormous

when compared to a single individual, which can measure as little as a few millimeters.

Scientists have been studying the coordination mechanisms that allow the construction of these

structures and have proposed probabilistic models exploiting insects behavior. Some of these

models are implemented in computer programs to produce simulated structures that recall the

morphology of the real nests.

c) FLOCKING AND SCHOOLING IN BIRDS AND FIFISH

Scientists have shown that these elegant swarm-level behaviors can be understood as the

result of a self-organized process where no leader is in charge and each individual bases its

movement decisions solely on locally available information: the distance, perceived speed, and

direction of movement of neighbours. These studies have inspired a number of computer

simulations that are now used in the computer graphics industry for the realistic reproduction of

flocking in movies and computer games.

d) ANT COLONY OPTIMIZATION

In ant colony optimization (ACO), 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 artificial ants incrementally build solutions by

moving on the graph. The solution construction process is stochastic and is biased by a

19

pheromone model, that is, a set of parameters associated with graph components the values of

which are modified at runtime by the ants.

e) PARTICLE SWARM OPTIMIZATION

It is inspired by social behaviors in flocks of birds and schools of fish. In practice, in the

initialization phase each particle is given a random initial position and an initial velocity. The

position of the particle represents a solution of the problem and has therefore a value, given by

the objective function. At each iteration of the algorithm, each particle moves with a velocity that

is a weighted sum of three components: the old velocity, a velocity component that drives the

particle towards the location in the search space where it previously found the best solution so

far, and a velocity component that drives the particle towards the location in the search space

where the neighbor particles found the best solution so far.

f) SWARM BASED NETWORK MANAGEMENT

Schoonderwoerd et al. proposed Ant-based Control (ABC), an algorithm for routing and

load balancing in circuit-switched networks; Di Caro and Dorigo proposed AntNet, an algorithm

for routing in packet-switched networks. While ABC was a proof-of-concept, AntNet, which is

an ACO algorithm, was compared to many state-of-the-art algorithms and its performance was

found to be competitive especially in situation of highly dynamic and stochastic data traffic as

can be observed in Internet-like networks. An extension of AntNet has been successfully applied

to ad-hoc networks.

g) COOPERATIVE BEHAVIOUR IN SWARMS OF ROBOTS

There are a number of swarm behaviors observed in natural systems that have inspired

innovative ways of solving problems by using swarms of robots. This is what is called swarm

20

robotics. In other words, swarm robotics is the application of swarm intelligence principles to the

control of swarms of robots. As with swarm intelligence systems in general, swarm robotics

systems can have either a scientific or an engineering flavour. Clustering in a swarm of robots

was mentioned above as an example of artificial/scientific system.

3.2.1 CURRENT PRACTICE IN SWARM INTELLIGENCE

The current and only practice of swarm intelligence is in the area of swarm robotics.

Swarm robotics is a new approach to the coordination of multi-robot systems which consist

of large numbers of mostly simple physical robots. It is supposed that a desired collective

behavior emerges from the interactions between the robots and interactions of robots with the

environment. This approach emerged on the field of artificial swarm intelligence, as well as

the biological studies of insects, ants and other fields in nature, where swarm behaviour

occurs.

The use of robots for tackling dangerous tasks is clearly appealing as it eliminates or reduces

risks for humans. The dangerous nature of these tasks implies a high risk of losing robots.

Therefore, a fault-tolerant approach is required, making dangerous tasks an ideal application

domain for robot swarms. Example of dangerous tasks that could be tackled using robot

swarms are demining, search and rescue, and cleaning up toxic spills.

Potential applications for robot swarms are those in which it is difficult or even impossible to

estimate in advance the resources needed to accomplish the task. For instance, allocating

resources to manage an oil leak can be very hard because it is often difficult to estimate the

oil output and to foresee its temporal evolution. In these cases, a solution is needed that is

21

scalable and flexible. A robot swarm could be an appealing solution: robots can be added or

removed in time to provide the appropriate amount of resources and meet the requirements of

the specific task. Example of tasks that might require an a priori unknown amount of

resources are search and rescue, tracking, and cleaning.

Another potential application domain for swarm robotics are tasks that have to be

accomplished in large or unstructured environments, in which there is no available

infrastructure that can be used to control the robots—e.g., no available communication

network or global localization system. Robot swarms could be employed for such

applications because they are able to act autonomously without the need of any infrastructure

or any form of external coordination. Examples of tasks in unstructured and large

environments are underwater or extraterrestrial planetary exploration, surveillance, demining,

and search and rescue.

Some environments might change rapidly over time. For instance, in a post earthquake

situation, buildings might collapse—thereby changing the layout of the environment and

creating new hazards. In these cases, it is necessary to adopt solutions that are flexible and

can react quickly to events. Swarm robotics could be used to develop flexible systems that

can rapidly adapt to new operating conditions. Example of tasks in environments that change

over time are patrolling, disaster recovery, and search and rescue.

3.3 ADVANTAGES AND DISADVANTAGES OF SWARM INTELLIGENCE

3.3.1 Benefits of swarm systems

22

Adaptable – Conventional workgroups devise various standard operating procedures to

react to predetermined stimuli. But swarms have better ability to adjust to new situations

or to change beyond a narrow range of options. Countless novel possibilities exist in the

exponential combinations of many interlinked individuals;

Evolvable – Evolution is the result of adaptation. Conventional bureaucratic systems can

shift the locus of adaptation (slowly) from one part of the system to another. In swarm

systems, individual variation and imperfection lead to perpetual novelty, which leads to

evolution;

Resilient – A swarm is a collective system made up of multitudes in parallel, which

results in enormous redundancy. Because the swarm is highly adaptable and evolves

quickly, failures tend to be minimal.

Boundless -- Plain old linear systems can sport positive feedback loops -- the screeching

disordered noise of PA microphone, for example. But in swarm systems, positive

feedback can lead to increasing order. By incrementally extending new structure beyond

the bounds of its initial state, a swarm can build its own scaffolding to build further

structure. Spontaneous order helps create more order. Life begets more life, wealth

creates more wealth, information breeds more information, all bursting the original

cradle. And with no bounds in sight.

Novelty -- Swarm systems generate novelty for three reasons:

(1) They are "sensitive to initial conditions" -- a scientific shorthand for saying that the

size of the effect is not proportional to the size of the cause -- so they can make a

surprising mountain out of a molehill.

23

(2) They hide countless novel possibilities in the exponential combinations of many

interlinked individuals.

(3) They don't reckon individuals, so therefore individual variation and imperfection can

be allowed. In swarm systems with heritability, individual variation and imperfection will

lead to perpetual novelty, or what we call evolution.

3.3.2 Disadvantages of swarm systems

Non-optimal – Because swarm systems are highly redundant and have no central control,

they tend to be inefficient. The allocation of resources is not efficient, and duplication of

effort is always rampant. Swarms can dampen inefficiency, but never to the degree that a

linear system can;

Uncontrollable – It is very difficult to exercise control over a swarm. Swarm systems

require guidance in the way that a shepherd drives a herd: by applying force at crucial

leverage points;

Unpredictable – The complexity of a swarm system leads to unforeseeable results.

Emergent novelty is a primary characteristic of self-organisation by adaptive systems.

Not all novelty is desirable;

Non-understandable – Sequential systems are understandable; complex adaptive systems,

instead, are a jumble of intersecting logic. Instead of A causing B, which in turn causes

C, A indirectly causes everything, and everything indirectly causes A;

Non-immediate – Linear systems tend to be very direct: Flip a switch and the light comes

on. Simple collective systems tend to operate simply. But complex swarm systems with

24

rich hierarchies take time. The more complex the swarm, the longer it takes to shift states.

Each hierarchical layer has to settle down, peripheral players have to come to rest, and a

multitude of autonomous agents need to become acquainted with each other.

CHAPTER FOUR

4.2 SUMMARY / CONCLUSION

The idea of swarm behavior may still seem strange because we are used to relatively linear

bureaucratic models. In fact, this kind of behavior characterizes natural systems ranging from

flocks of birds to schools of fish. Humans are more complex than ants or fish and have lots more

capacity for novel behavior, some unexpected results are likely, and for this reason, leading

scientists and organizations will further pursue swarm approaches. Swarm Intelligence provides

a distributive approach to the problem solving mimicking the very simple natural process of

cooperation. According to my survey many solutions that had been previously solved using other

AI approach like genetic algorithm neural network are also solve able by this approach also. Due

to its simple architecture and adaptive nature like ACO has it is more likely to be seen much

more in the future.

25

L. Bianchi, L. M. Gambardella, and M. Dorigo. An ant colony optimization approach to

the probabilistic traveling salesman problem. In J. J. Merelo,

P. Adamidis, H.-G. Beyer, J.-L. Fern´andez-Villacanas, and H.-P. Schwefel, editors,

Proceedings of PPSN VII, Seventh International Conference on Parallel Problem Solving

from Nature, volume 2439 of Lecture Notes in Computer Science, pages 883–892.

Springer, Berlin, Germany, 2002.

B. Bilchev and I. C. Parmee. The ant colony metaphor for searching continuous design

spaces. In T. C. Fogarty, editor, Proceedings of the AISB Workshop on Evolutionary

Computation, volume 993 of Lecture Notes in Computer Science, pages 25–39. Springer,

Berlin, Germany, 1995.

M. Birattari, G. Di Caro, and M. Dorigo. Toward the formal foundation of ant

programming. In M. Dorigo, G. Di Caro, and M. Sampels, editors, Ant Algorithms

Proceedings of ANTS 2002 – Third International Workshop, volume 2463 of Lecture

Notes in Computer Science, pages 188–201. Springer, Berlin, Germany, 2002.

S. Bird and X. Li. Adaptively choosing niching parameters in a PSO. In Mike Cattolico,

editor, Genetic and Evolutionary Computation Conference, GECCO 2006, Proceedings,

Seattle, Washington, USA, July 8-12, 2006, pages 3–10. ACM, 2006..

26

T. Blackwell and P. J. Bentley. Dynamic search with charged swarms. In Proc. the

Workshop on Evolutionary Algorithms Dynamic Optimization Problems (EvoDOP

2003), pages 19–26, 2002.

T. Blackwell and P. J. Bentley. Improvised music with swarms. In David B.Fogel,

Mohamed A. El-Sharkawi, Xin Yao, Garry Greenwood, Hitoshi Iba, Paul Marrow, and

Mark Shackleton, editors, Proceedings of the 2002 Congress on Evolutionary

Computation CEC 2002, pages 1462–1467. IEEE Press, 2002.

T. Blackwell and J. Branke. Multi-swarm optimization in dynamic environments. In

EvoWorkshops, volume 3005 of Lecture Notes in Computer Science, pages 489–500.

Springer, 2004. T. Blackwell and J. Branke. Multi-swarms, exclusion and anti-

convergence in dynamic environments. IEEE Transactions on Evolutionary Computation,

10(4):459–472, 2006.

M. Blesa and C. Blum. Ant colony optimization for the maximum edge-disjoint paths

problem. In G. R. Raidl, S. Cagnoni, J. Branke, D. W. Corne, R. Drechsler, Y. Jin, C. G.

Johnson, P. Machado, E. Marchiori, R. Rothlauf, G. D. Smith, and G. Squillero, editors,

Applications of Evolutionary Computing, Proceedings of EvoWorkshops 2004, volume

3005 of Lecture Notes in Computer Science, pages 160–169. Springer, Berlin, Germany,

2004.

C. Blum. Beam-ACO Hybridizing ant colony optimization with beam search: An

application to open shop scheduling. Computers & Operations Research, 32(6):1565–

1591, 2005.

27