SWARM ROBOTICS

CONTENTS:

 Acknowledgement
 Overview: swarm intelligence
 The Boid Model
 Origin of swarm intelligence
 Swarm intelligence to combat marine oil pollution
 Application in disaster control
 Future of swarm robotics
 Latest developments in the field
 Bibliography
 SWARM ROBOTICS
“Swarm intelligence works best when all parts are equally stupid!!!!”

 What if computers could make decisions for themselves?

 What if you dropped a batch of computer parts onto a deserted island and they
built a civilized communications network all by themselves.

 What if your doctor injected a swarm of microchips to attack a hard-to-reach

tumor?

A few years ago, those were very big ifs.

Sounds scary?

Well, there's no need for getting scared. We're not talking artificial intelligence -- just
swarm intelligence.
To find the effective solutions for the everyday problems, researchers are now
increasingly concentrating on swarm intelligence based on the collective strength of fish,
birds and insects.
As a single system, these animals can together hit upon a solution for complicated tasks.
We are introducing a new concept that is still at its initial level of development but is one
of the most promising field in the coming future.
As well we introduce some fields where this concept is been already employed or can be
employed successfully as more and more advances are made in the practical aspect.

OVERVIEW: SWARM INTELLIGENCE

As the name implies, swarm intelligence is a scientific theory based on the actions of
ants, bees and other insects.
It asserts complex behavior can emerge from a group of individuals -- whether they're
bugs or computer nodes -- that follow simple rules.
SI is continuously incorporating new ideas, algorithms, and principles from the
engineering and basic science communities.
Application of the swarm principles to a large number of robots is known as
‘SWARM ROBOTICS’.

THE BOID MODEL:

A collective group manages well even with a few rules.
For instance, the flight of a flock of birds can be described in three patterns:
First claws out before the collision. second, fly as fast as the neighbor and third, fly in the
direction perceived by other birds.
In computer simulation, these simple rules are adequate to replicate realistic group
behavior.
Animation studios in Hollywood are using this principle to create collective groups of
animals tat remarkably behave like their real counterparts.
In 1987, Reynolds created a boid" model, which is a distributed behavioral model, to
simulate on a computer the motion of a flock of birds .Each boid is implemented as an
independent actor that navigates according to its own perception of the dynamic
environment.
A boid must observe the following rules.
First, the “avoidance rule” says that a boid must move away from boids that are too
close, so as to reduce the chance of in-air collisions.
Second, the “copy rule”says a boid must fly in the general direction that the flock is
moving by averaging the other boids' velocities and directions.
Third, the “center rule” says that a boid should minimize exposure to the flock's exterior
by moving toward the perceived center of the flock.
fourth rule “view rule” that indicates that a boid should move laterally away from any
boid the blocks its view.
This boid model seems reasonable if we consider it from another point of view, that of it
acting according to attraction and repulsion between neighbors in a flock. The repulsion
relationship results in the avoidance of collisions and attraction makes the flock keep
shape, i.e., copying movements of neighbors can be seen as a kind of attraction. The
center rule plays a role in both attraction and repulsion.
The swarm behavior of the simulated flock is the result of the dense interaction of the
relatively simple behaviors of the individual boids.
Although this scenario seems to be well under-stood in biology, the mechanisms
underlying cooperative transport remain unclear. Roboticists have attempted to model
this cooperative transport.

For instance, Cube and Zhang [2] introduce a simulation model including stagnation
recovery with the method of task modeling
SWARM INTELLIGENCE : INTRODUCTION & ORIGIN

The ants find the shortest route. Thanks to their predecessors. If the routes are similar,
then all take the one round first. Computer scientists are already applying this tactic in
many cases.

Algorithms based on animal behavioral patterns:

The manner in which the ants use pheromone, a scent marking to track the shortest way
between the source of food and dwelling serves as a model for finding solutions to
problems based on logistics and optimization Such as in case of routing in
telecommunication networks.

The software that is based on swarm of ants replicates the behavioral pattern of its role
model by gradually establishing the variables of a solution that is sought after-
resembling ants searching their way.
 A pressing task of the software is to record the so called context awareness i.e. using the
available data to find out the state in which an individual or object is and how an
intelligent environment can react.

Siemens is already applying this technology in the fields of home networking and
building engineering services.

Explaining the principle behind this technology, Dr Cornel Klein from the Siemens
Research Department of Corporate Technology said,” A common software architecture
serves as a universal link for the solutions of various problems.

SWARM TO COMBAT MARINE OIL POLLUTION!!!!!!!!

A project for swarm robotics is already taking a shape in the sea.

Researchers at the Fraunhofer Institute for the Manufacturing Engineering and

Automation (IPA), Germany are geared toward using this technology to control oil spill.
The robots in this case are the small ships resembling catamarans with an exhaust device
between the vats.

Dennis Fritsch, robot expert at IPA is of the opinion that a swarm with distributed
intelligence is perfectly suited in the field of oil. According to him it is not possible to
have a central planning system since the prediction of the nature of an oil spill is
unreliable.

Thus, 100 robots after taking their respective positions, communicate with one another,
find oil and then absorb it.”

Swarm intelligence works best when all parts are equally stupid”. Specialization within
the swarm has only negative implications on the overall result. Moreover failures can be
better compensated if all ‘employees’ are identical.
Fritsch says that as long as sensor communicates its presence in the oil spill to the robot,
the process is smooth. The robot rotates coils until it discovers the oil.

The number of the robots is directly proportionate to how well the waste can be
eliminated.

Currently this method is being used in the simulation with small mobile robots. A beamer
at the base indicates an expanding oil spill. As a part of the EU project, it will be decided
whether the entire plan can be actually put into practice. If the decision is positive then
the first robotic fleets will be deployed in the next five years.

As can be seen in the figure on the left:

1 A swarm of robots are set from the ship into the water in the direction of the oil
spill.
2 As long as the sensors measure the oil, the robots simply continue going ahead.
3 If the robots come out of the oil spill, they rotate coils until the sensors detect the
oil again.
4 The robots go back to the ship carrying the oil waste.
Lets see the concept in sensor network: A sensor network displays the temperature, the
centre of smoke and pollutants in its environment to fire fighters and guides them to the
source of fire.
APPLICATION IN DISASTER CONTROL:

DIKES AND WINDMILLS DO THE THINKING!!!!!!!!!!

At the German coast of the Baltic Sea, researchers are working on using swarm
computing for the disaster control.
Project swarm developed by the institute for Telematics of Lubeck University of
communication Systems at Kiel involves wireless networking of sensors systems that can
protect dikes and coasts.
Away from the sand bags at the dike, a swarm with modules will be set up after every
meter, and will be provided with a small computer, GPS, various sensors and a small
radio interface.
If the sensors record a water inrush, then this information is immediately forwarded
Presently, a major hurdle of this project is the exact location of the swarm.
A sand sack situated deep in the seawaters has no visibility and thus will hardly be able to
receive a GPS signal. Researchers of the European Galileo project are working on a way
to make it possible to receive stronger signals in the coming years.
While sensors in the dike ring the alarm bell in case of danger and allow the emergency
measures to kick off, other sensor networks are directly linked with a control system.
Offshore wind parks would be able to provide more energy and would be less sensitive
had they been able to communicate with one another.
In big windmills such as those already seen on the Danish coasts, thousands of rotors set
up in several rows are arranged at a stretch. Siemens has connected these windmills in
such a way that they mutually monitor one another
The windmills in the front rows send updated measured wind data to the mills arranged
at the back. These mills can then immediately adapt to the blowing wind if change the
alignment of the rotor blades in case of dangerous gusts.
What makes these ‘futuristic creature’ so fascinating?
It is a fact that although these creatures exchange information, an individual creature need
not be super bright; they are programmed to work in a swarm, which in turn results in
intelligence.
These machines are installed with GPS cameras, and image recognition systems for
the orientation.
A collective group manages well even with a few rules.
Social insects are known to coordinate their actions to accomplish tasks that are beyond
the capabilities of a single individual:
Termites build large and complex mounds, army ants organize impressive foraging
raids, ants can collectively carry large preys.
Such coordination capabilities are still beyond the reach of current multi-robot systems.
Networks resembling a swarm could play a more important role for energy supply in
future. If in future, in a planned European integrated network, more power supply is
supplied from renewable energy sources such as water, wind and sun, then these gigantic
networks can no longer be regulated from a control center.”
“We need to think of how the power supply can controls itself”, says Professor Martin
Greiner from Siemens.
The Professor’s vision is to distribute the intelligence within the power supply network
in order to immediately pick up as many fluctuations as possible in the nodal points
and not later in already overloaded control center. In this way blackouts can be
prevented
The latest developments:
 Distributed intelligence is already being used in the road traffic. Pilot projects in
congested urban areas combine central traffic management system with end
devices such as navigation system and cell phones. Thousands of sensors collect
data for this purpose. Hidden in paving, at bridges or poles, these sensors collect
information through traffic density as well as the direction and speed of the
vehicles.
 In accident prone areas like a tunnel, it is very important to have an automatic
analysis of data available on the site. A good example is that of the Swiss Giswil
tunnel that has an intelligent video monitoring system.
The entire set up as can be seen in the figure on left can not only detect fire with
help of smoke emission at an early stage but also identify stagnated vehicles or
even a traffic jam.
For this purpose, the evaluation algorithm compares the current picture with the
reference pictures already archived.
In case of the persistent change, the system triggers an alarm and sets off a chain
reaction of the security measures, for instance, traffic lights automatically switch
to red and the smoke outlet is activated.
 Having drawn inspiration from ants, researchers at Siemens, Germany have
envisioned small mobile machines that will replicate the behavior of giant
ants.
 At a skyscraper in New York, these machines pick up information from the
environment, communicate with one another and together solve a problem
assigned to them.
 We have seen it in movies like Minority Report where a throng of digitally
controlled artificial ants track the movements of the central character played by
Tom Cruise.
Experts are hopeful that such autonomous network systems could one day inspect
and carry out maintenance work in case of dangerous situations or even take over
mobile monitoring tasks
CONCLUSION

We have introduced a new concept that is still at its initial level of development but is
one of the most promising field in the coming future.
We introduced some areas where in this concept is been already employed or can be
employed successfully as imagination meets technology.
BIBLIOGRAPHY

 Intelligent computing chip-february 2007

 Minority report movie.