The Power of Simplicity

Presented By Parvinder Singh Palak Bhatia Laxmi Suri

What is Swarm Intelligence (SI)? The emergent collective intelligence of groups of simple agents. Swarm intelligence (SI) as defined by Bonabeau, Dorigo and Theraulaz is any attempt to design algorithms or distributed problem-solving devices inspired by the collective behavior of social insect colonies and other animal societies

An artificial intelligence (AI) technique based on the collective behavior in decentralized, self-organized systems Generally made up of agents who interact with each other and the environment No centralized control structures Based on group behavior found in nature

Swarm Intelligence

Examples

Swarms

build colonies and work in a coordinated manner yet no single member of the swarm is in control. Flocks of birds coordinate to move without collision. Ants manage to find food sources quickly and efficiently. Termites build giant structures. Schools of fish fend off predators and move as one body

SI - The Beginnings
First

introduced by Beni and Wang in 1989 with their study of cellular robotic systems

The

concept of SI was expanded by Bonabeau, Dorigo, and Theraulaz in 1999 (and is widely recognized by their colleges)

Why do we need new computing techniques?

The computer revolution changed human societies:

Communication & Transportation Industrial production Administration, writing and

bookkeeping Technological advances Entertainment

However, some problems cannot be tackled with traditional hardware and software!

Drawbacks of traditional techniques

Computing tasks have to be :Well-defined Fairly

predictable Computable in reasonable time with serial computers.

Working

Two principles of Swarm Intelligence

Self-organization Stigmergy

- stimulation

by work

Why Social Colony is a source of inspiration?

Flexible:

internal challenges Robust: tasks are completed even if some individuals fail Decentralized: there is no central control(ler) in the colony Self-organized: paths to solutions are emergent rather than predefined

the colony can respond to perturbations and external

Ants
Why are ants interesting?
ants solve complex tasks by simple local means ant productivity is better than the sum of their single activities ants are grand masters in search and exploitation

Which mechanisms are important?

cooperation and division of labour adaptive task allocation work stimulation by cultivation

Pheromone Trails
Species

lay chemical substance pheromone while travelling from nest, to nest or possibly in both directions. evaporate.

Pheromones Pheromones

accumulate with multiple ants using same path.

Ant Colony Optimization (ACO)

Is

inspired by the behavior of ant colonies . Ability of Optimization in finding shortest path. Ants leave a chemical pheromone trail. Pheromone trails enables them to find shortest paths between their nest and food sources Ants find the shorter path in an experimental setup

bridge leads from a nest to a foraging area, (a) 4 minutes after bridge placement, (b) 8 minutes after bridge placement

Ant Foraging
Cooperative search by pheromone trails

1. The natural behavior of these ants and be programmed into an ant algorithm, which we can use to find the shortest path within graphs. 2. As ants move they leave behind a chemical substance called pheromone, which other ants can smell and identify that an ant has been there before.

ACO algorithm
Main steps of the ACO algorithm are given below: Pheromone trail initialization Solution construction using pheromone trail Each ant constructs a complete solution to the problem according to a probabilistic State transition rule. The state transition rule depends mainly on the state of the pheromone . Pheromone trail update.

Applications of Ant Colony Optimization

Traffic on telecommunications systems, the internet, roads, rail, and sea would all benefit from the reduction in congestion that efficient routing algorithms could provide. Modern airlines are actually putting the ant colony research to work, with impressive payback

Telecommunication System

Airlines

Cont.
Vehicle

routing with time window constraints Network routing problems Assembly line balancing Data mining

Swarm Intelligence

Particle Swarm Optimization (PSO)

Idea:

Used to optimize continuous functions

Function

is evaluated at each time step for the agents current position

 Each

agent remembers personal best value of the function (pbest) Globally best personal value is known (gbest) Both points are attracting the agent

Solving some of the NPHard Problems using Swarm Intelligence

1. Well-defined, but computational hard problems

NP hard problems (Travelling Salesman Problem) Action-response planning (Chess playing)

2. Fuzzy problems intelligent human-machine interaction natural language understanding

3. Hardly predictable and dynamic problems real-world autonomous robots management and business planning

Example of such Problems are:

Complex

NP complete problems. Vehicle routing. Network maintenance. The traveling salesperson. Computing the shortest route between two points.

Travelling Salesman Problem

1.

Visit cities in order to make sales. 2. Save on travel costs. 3. Visit each city once (Hamiltonian circuit).

Solution of TSP by Swarm Intelligence

Use

of agents for TSP problem. They sense and dispense pheromone. Memory to back step through the graph. Each agent starts at a random starting city. Once agent finishes a tour, it determines the size of the tour. Then pheromone is added to the tour, the shorter the tour, the higher

 No guarantee that the first tour the

agents will converge the shortest path. Agents explore other tours. The stray agent finds a shorter path. Adjusts the pheromone levels. Plenty of computing time needed to converge on the optimal tour. The ant algorithm approach will still solve faster than other

Vehicle routing by Swarm Optimization

Vehicle

routing is similar to the TSP problem. Employee services the client by going to them. Minimize cost. Use the same optimal Hamiltonian circuit as in the TSP problem.

Use of Swarm Intelligence in Economy

The

economy is an example of SI that most researchers forget to consider. SI demonstrates complex behavior that arises from simple individual interactions. No one can control the economy, as there are no groups that can consistently

The

reaction of the population causes the economy to slow down. an economy using ant

Simulating

algorithms.
Swarm

intelligence, is still in its infancy (growth).

Why is Swarm Intelligence interesting for IT? Analogies in IT and social insects distributed system of interacting

autonomous agents goals: performance optimization and robustness self-organized control and cooperation (decentralized) division of labor and distributed task allocation indirect interactions

How do we design Swarm Intelligence Systems?

It is a 3-step process.
1.Identification

of analogies: in swarm biology and IT systems. 2.Understanding: computer modeling of realistic swarm biology. 3.Engineering: model simplification and tuning for IT applications.

Pros and Cons

Cons Difficult to predict collective behaviour from individual rules. Interrogate one of the participants, it wont tell you anything about the function of the group. Small changes in rules lead to different group-level behaviour. Individual behaviour looks like noise: how do you detect threats? Pros Possible to efficiently control organization or manipulate groups using simple rules. Possible to predict group-level outcome using bottom

Applications of SI
Swarm/crowd

simulation programming Computer Networks: Adaptive Routing Robotics/Artificial Intelligence Process optimization /Staff Scheduling

Conclusion
Scientists The

are realizing SIs potential.

use of ant algorithms within computing systems has helped to solidify swarm intelligences place in the computing world. researchers are observing other social animals, such as bees and schools of fish in order to utilize it in future applications and algorithms.

Already

References
http://www.engr.iupui.edu/~eberhart

http://users.erols.com/cathyk/jimk.ht

ml

http://www.alife.org/ http://www.aridolan.com/

Any Questions?