Swarm Robotics: A Developing Field

In our nature, we get to see a lot of wonderful team work, even among very poorly evolved species. Examples include ant colonies, bird flocking, animal herding, bacterial growth, and fish schooling. Here, instead of individual intelligence, these creatures show a collective behaviour known as swarm behaviour and thus possess what can be called swarm intelligence.  Key features of such behaviour are that each individual follow simple rules and there is no centralised control structure dictating their behaviour but the seemingly random and localised interaction between such individuals lead to the emergence of “intelligent” global behaviour.  Such behaviour has great scope in the field of robotics.

 

The artificial simulation of swarm behaviour can be traced back to 1986, when an artificial life and computer graphics expert Craig Reynolds developed a program called Boids. It could simulate the flocking behaviour of birds. Simple rules like separation, alignment and cohesion were able to simulate emergent behaviour where complex systems and patterns arise out of a multiplicity of relatively simple interactions. This program has seen many interesting uses such as to create realistic-looking representations of flocks of birds and other creatures. Examples include flying bird-like creatures in the famous video-game Half-Life and swarm of bats in the Batman Returns feature film.

 

In a swarm, each individual element is called an agent and will be an autonomous entity capable of observing and acting upon an environment and directing its activity towards achieving goals. Swarm robotics is a new approach to the coordination of multi-robot systems where a desired collective behaviour emerges from the interactions between the robots and interactions of robots with the environment. Generally swarm robotics differ from distributed robotic systems in the aspect that it emphasizes large number of robots, and promotes scalability. Scalability is achieved by using localised communication via wireless transmission systems, like radio frequency or infrared.  A swarm-intelligent approach tries to achieve meaningful behaviour at swarm-level, instead of the individual level.

 

One of the most interesting advancements in the field of swarm robotics was by the project team headed by Dr. Marco Dorigo with their work “Swarm-bots: Swarms of self-assembling artifacts”. It was aimed to study new approaches to the design and implementation of self-organizing and self-assembling artifacts. They designed small, autonomous mobile robots called s-bots which are capable of performing basic tasks such as autonomous navigation, perception of its surrounding environment, and grasping of objects. A self-assembling and self-organising robot colony made of a number (30-35) of s-bots is called a swarm-bot and is capable of performing exploration, navigation and transportation of heavy objects on very rough terrains, especially when a single s-bot has difficulty in achieving the task alone. This project was limited to a 2-dimensional terrain. It lasted 42 months, was successfully completed on March 31, 2005.

 

The Swarmanoid project is extending the work done in the Swarm-bots project to three dimensional environments. This system is made up of heterogeneous, dynamically connected, small autonomous robots, collectively called a swarmanoid. It comprises of three robot types: eye-bots, hand-bots, and foot-bots. The eye-bots are able to fly and attach to the ceiling and to sense and analyse the environment from a high position. The hand-bots can climb vertical surfaces of walls or other objects. Foot-bots are specialised in moving on rough terrain and transporting either objects or other robots (based on s-bots). Together these 3 types of agents form a heterogeneous system capable of moving in 3-dimensional space. This project completed successfully on September 30, 2010. A very impressive video titled “Swarmanoid, the movie” describing the project and demonstrating an example where these robots together steal a book from a shelf is available in the Internet. It also won the AAAI 2011 video competition.

 

The potential application of swarm robotics are in semi-automatic space exploration, search for rescue, underwater exploration etc. These can also be used to perform tasks that demand for miniaturization (nanorobotics, microbotics), like distributed sensing tasks in micromachinery or the human body, tasks that demand cheap designs (mining tasks or agricultural foraging tasks) etc. A study of the artificial swarm behaviour also helps us obtain a better understand biological swarming (bird and insect migration, bee and ant colonies, fish shoaling and schooling etc). Thus the possibilities offered by this developing field of robotics are immense.

 [highlight color=”yellow”]Rahul.R (CET)[/highlight].