Digital Ant Colony System

Whoever thinks that ants are only erratic little beings, whose incessant wanderings are pointless or obey no kind of universal order, is very much mistaken. Their society is so organized and complex that it is already being used as a model for the creation of algorithms, adaptable to a variety of scientific areas.

Inspired by the work of Chialvo and Millonas – the creators of the first numeric simulation according to which the formation of collective cognitive maps of social insects can be explained -, Vitorino Castelo Ramos, a young Portuguese researcher in the area of Artificial Research and Image Analysis, has developed a computational model that has a direct application to habitats of digital images.

More than simply reacting to the environment presented to them (any image digitally inserted as substratum), the digital ants of these colonies develop the capacities of sensing objects and outlines just like real ants do, adapting themselves to any digital environment.

After publishing the scientific work about the application of the “Ant colonies Systems” for the classification of databases and document clustering in February 2002, Vitorino Ramos has continued to dedicate himself to the development of algorithms that allow the practical application of this model.

An example of this are the algorithms that imitate the Stigmergy phenomenon – or rather, a certain environmental synergy that translates into the organization of ant colonies in different spatial groups – like the behaviour of real ants, with Data Mining applications.

Apart from serving as a base for the notion of the first operators of completely distributed, adaptable and self-organized Mathematical Morphology, the “Ant colony System” has many other possible applications, such as on an architectural and urban planning level or also for painting done by robots.