The Principles of Self Organization - Stigmergy
by Josh Patterson ~ August 25th, 2008. Filed under: Self Organization.(This article is part of an ongoing series “The Principles of Self Organization“, preceeded by the article “Negative Feedback“.)
Stigmergy is a method of communication, albeit in an indirect fashion, between two units or agents. Stigmergy is defined as information gathered from work in progress [ 2 ]. It is discussed with respect to self organization, an approach that emphasizes:
- direct or indirect interactions among relatively simple agents
- distributedness
- flexibility
- robustness
Social insects are a primary research area with regards to stigmergy and self organization. Self organization in social insects often requires interactions among insects that can be direct or indirect. A direct interaction would be contact between two insects such as antennation or trophallaxis (liquid or food exchange). Indirect interactions are more latent, however, as they occur when one individual modifies the environment which another individual insect or agent responds to at another time. We are mostly interested in the indirect type of interaction here as this is what we call “stigmergy”.
Grasse [ 3, 4 ] introduced stigmergy (from the Greek stigma: sting, and ergon: work) to explain task coordination and regulation in the context of nest reconstruction in termites. He showed that the building activities of these insects do not depend on the workers themselves but on the current state of the nest itself; any given local state of nest construction triggers a response in a worker which in turn modifies the local state to the next step in its “rule set”. This continues until a final or stable state of the nest is achieved. This type of decentralized construction has been simulated [ 1 ] a number of times with very interesting results.
For social insects to rely on information derived from the environment as opposed to other local individuals in their colony or a central leader may seem perplexing at first. However, there is a good reason for the employment of stigmergy as it allows for scalable decentralized coordination amongst relatively unsophisticated agents. Top down control hierarchies tend to lose the ability to efficiently manage large groups of agents whereas self organizing processes scale with much success. Social insects are able to operate with no central leader and exhibit the emergent collective intelligence of groups of simple agents.
As our world becomes more complex with a developing internet as a platform with an evolving ecosystem of linked data, we face many obstacles with regards to making sense of the complexity of the world around us. Business, Government, and software systems are producing more data than ever, becoming so intractable that they can no longer be controlled in a centralized fashion.
Swarm intelligence, or self organization, is one method that offers an alternate way of designing intelligent systems, system which exhibit autonomy, emergence, and distributed functionality. These properties can augment or even replace systems that employ techniques such as control, preprogramming, and centralization. Systems which employ stigmergy as an operational property are flexible and adaptive. These systems can respond to environmental changes as if the changes were inherently part of the system’s operations and continue to not only operate, but thrive. Stigmergy is a key computational component of most self organizing systems, which is why I illustrate its properties in this article as a building block to study and reference later on.
References
[1] E. Bonabeau, M. Dorigo, G. Theraluz, Swarm Intelligence: From Natural to Artificial Systems, Oxford University Press, 1999.
[2] S. Camazine, J.L. Deneubourg, N. Franks, J. Sneyd, G. Theraluz, E. Bonabeau, Self Organization in Biological Systems, Princeton University Press, 2001
[3] Grasse’, P.-P. “La Reconstruction du nid et les Coordinations Inter-Individuelles chez Bellicositermes Natalensis et Cubitermes sp. La Theorie de la Stigmergie: Essai d’interpretation du Comportement des Termites Constructeurs.” Insect. Soc. 6 (1959): 41-80
[4] Grasse’, P.-P. “Termitologia, Tome II.” Fondation des Societes. Construction. Paris: Masson, 1984
September 12th, 2008 at 2:57 pm
[...] Organization. The first article was Emergence, and the article that preceeded this one was based on stigmergy. ) What is it that governs here? What is it that issues orders, forsees the future, elaborates [...]
November 23rd, 2008 at 11:28 pm
[...] inspired routing algorithm for MANETs that is based on previous work in adapting the effects of stigmergy. Termite models how termite colonies lay pheromone to communication in a decentralized manner. The [...]