SRSP

• Definition
• Characteristics
• Desired properties
• Domains of application
• Research axes
  • Design
  • Modeling
  • Collective Behaviors
• Current limits

---

OLD

Swarm robotics studies how a large number of embodied agents can be designed and controlled so that a desired collective behavior results from local interactions among the agents and between the agents and the environment in which they act (Dorigo and Sahin 2004; Sahin 2005). The design of swarm robotics systems is guided by swarm intelligence principles. The main goal pursued is the development of physical swarms exhibiting collective behaviors that are robust, scalable and flexible.

What Characterizes Swarm Robotics

Research in swarm robotics focuses on robotics systems characterized by the following properties:

• the robotics system consists of a large group of autonomous robots, called a swarm;
• the robots in the swarm are relatively homogeneous (i.e., they are either all identical or they belong to a few typologies);
• the interactions among the robots in the swarm are based on simple behavioral rules that exploit only local information that the robots exchange directly or via the environment (stigmergy);
• the overall behavior of the system results from the interactions of the robots with each other and with their environment, that is, the swarm behavior self-organizes;
• the robots in the swarm are relatively incapable with respect to the tasks that they are asked to perform;
• physical and/or logical cooperation of the robots allows the swarm to overcome the limitations of the single robots.

Desirable properties of swarm robotics systems

Some of the most desirable properties that are searched for in a swarm robotics system are scalability, robustness and fault tolerance. Important design choices that help in achieving these system level properties are the above mentioned use of local sensing and communication and of distributed control.

Fault tolerance

Because in swarm robotics there are many instances of each robot type, and because the control is typically decentralized, the failure of one or more individuals does not necessarily determine the failure of the system as a whole. In general, in a properly designed swarm robotics system failures are associated to either no change in performance, when the number of available robots is higher than the number of robots necessary for performing the considered task, or to a graceful degradation of the system performance. Additionally, the relative simplicity of the robots, in their sensing/control/actuation aspects, make fault less probable than in more complex robots.

Flexibility

By flexibility we mean that the swarm robotic system should have the ability to tackle different task by only changing the coordination strategy. Ants provide impressive examples of flexibility. On one hand, during foraging, ants move independently of each other and coordinate their search through pheromones laid on the ground. On the other hand, prey retrieval task requires the ants to generate a force much larger than that of a single individual to drag a prey to the nest. When a large prey is discovered, each ant grip the prey with its mandible and pull it in different directions. The seemingly random pulls of ants are observed to be coordinated through the force integrated over the prey. Swarm robotic systems should also have the flexibility to offer solutions to the tasks at hand by utilizing different coordination strategies in response to the changes in the environment.

Scalability

By scalability, we mean that the swarm robotic system should be able to operate under a wide range of group sizes. That is, the coordination mechanisms that ensure the operation of the swarm should be relatively undisturbed by changes in the group sizes. This requirement implicitly poses the local interaction and communication constraint for coordination. Stigmergy, that is the use of environment as a medium of interaction, naturally leads towards scalable coordination methods.

Domains of application

The application of swarm robotics methods to real-world problems will depend on the mass production technologies of robots. Advances in mechatronics technology have already started shrinking the size and costs of autonomous robots\cite{iSwarm}. MEMS (Micro-Electro-Mechanical System)technology has started to make impressive progress towards the integration of sensing, actuation and computation on silicon substrate paving way to the mass production of micro and nano robots.

We believe that swarm robotics could be used for many real-world tasks. However, instead of giving specific examples, we will discuss what types of tasks swarm robotics approach would be most appropriate.

Tasks that cover a region or space

Swarm robotic systems are completely distributed and would be well suited for tasks that are concerned with the monitoring of a region or space. Environmental monitoring, or even cleaning would constitute a good domain of application.

Tasks that are too dangerous

In swarm robotic systems, individuals are dispensible and the performance of the system degrades gracefully even when it loses its members. Hence tasks such as de-mining a mine field or the surveillance of a battlefield from sky would be better addressed by swarm robotic systems.

Tasks that scale-up or scale-down in time

Swarm robotic systems rely on coordination mechanisms that are relatively independent of the swarm size. Hence, it allows the user to scale-up or scale-down the size the swarm based on the size of the task.

Tasks that require redundancy

Swarm robotic systems are inherently redundant, since the loss of an individual can be automatically compensated by others. Such a redundancy would make swarm robotic systems less prone to catastrophic failures.

Research axes

The research in swarm robotics can be roughly categorized into four broad categories:

Design

Most of the studies on swarm robotic systems are concerned with the development of coordination methods to solve certain tasks. Some of these studies are concerned with principled/automatized approaches to develop such behaviors, such as evolutionary approaches, whereas others are concerned with the manual development of behaviors to address certain tasks.

Modelling and Analysis

Swarm robotics systems, by their very nature, are difficult to control rely on self-organization. Hence, whereever possible modelling and analysis of swarm robotic systems can provide guarantees regarding the system-level properties.

Tools

As with any other field, swarm robotics depends highly on advances in both experimental hardware and simulation tools that aid the development of swarm robotic systems.

Applications

Finally, there are a few studies that deal with the actual or proposed applications of swarm robotics to real-world problems.

References

E. Bonabeau, M. Dorigo, and G. Theraulaz. Swarm Intelligence: From Natural to Artificial System. Oxford University Press, New York, 1999.

J.-L. Deneubourg, S. Aron, S. Goss, and J.-M. Pasteels. The self-organizing exploratory pattern of the Argentine ant. Journal of Insect Behavior, 3:159–168, 1990.

G. Di Caro and M. Dorigo. AntNet: Distributed stigmergetic control for communications networks. Journal of Artificial Intelligence Research, 9:317–365, 1998.

M. Dorigo, V. Maniezzo, and A. Colorni. Positive feedback as a search strategy. Technical Report 91-016, Dipartimento di Elettronica, Politecnico di Milano, Milan, Italy, 1991. Revised version published as: M. Dorigo, V. Maniezzo, and A. Colorni. Ant System: Optimization by a colony of cooperating agents. IEEE Transactions on Systems, Man, and Cybernetics – Part B, 26(1):29–41, 1996.

M. Dorigo and E. Şahin (Eds.). Special Issue on Swarm Robotics. Autonomous Robots, 17:111–246.

M. Dorigo and T. Stützle. Ant Colony Optimization. MIT Press, Cambridge, MA, 2004.

J. Kennedy and R. C. Eberhart. Particle swarm optimization. Proceedings of IEEE International Conference on Neural Networks, IEEE Press, Piscataway, NJ, pp. 1942-1948, 1995.

J. Kennedy, R. C. Eberhart, and Y. Shi. Swarm Intelligence. Morgan Kaufmann, San Francisco, CA, 2001.

E. Lumer and B. Faieta. Diversity and adaptation in populations of clustering ants. Proceedings of the Third International Conference on Simulation of Adaptive Behavior: From Animals to Animats 3, MIT Press, Cambridge, CA, pp. 501-508, 1994.

R. Schoonderwoerd, O. Holland, J. Bruten and L. Rothkrantz. Ant-based Load Balancing in Telecommunications Networks. Adaptive Behavior, 5(2):169–207, 1996.

External Links

• www.aco-metaheuristic.org: These are the official web pages dedicated to collect information about ACO.
• www.swarmintelligence.org: These are the official web pages dedicated to collect information about PSO.
• Swarm Intelligence: The main journal in the field.
• ANTS - From Ant Colonies to Artificial Ants: A Series of International Workshops on Ant Algorithms: This series of conferences, held for the first time in 1998, is the oldest in the swarm intelligence field.
• IEEE Swarm Intelligence Symposia: Another more recent, started in 2003, series of conferences dedicated to swarm intelligence.