An XSet Based Protocol for Coordinating the Behaviour of Stigmergic Ant-Like Robotic Devices

Abstract

Formal protocols for driving swarms of robotic devices towards predictable emergent behaviour are rare in the literature but critical in most upcoming sciences such as nanotechnology, genetic engineering, and amorphous computing. Such collections of robotic device control protocols would foster the development of deliberately engineered swarm outcomes, while at the same time, discouraging the feared evolution of "engines of destruction to nature and mankind". Many inspiring swarm coordination paradigms have been proposed in the literature. However simulated ant inspired metaphors have demonstrated plausible successes in terms of robustness, fault tolerance, speed, and quality of the emergent behaviour thereof. We propose an XSet based protocol with which swarms of stigmergic ant-like robotic devices are driven towards predictable shortest path finding and path following behaviours. The theme features of the proposed XSet based protocol are built on the primitive activities of simulated stigmergic foraging ant agents. In this context, an XSet is an extended set of primitive activities, parameter values, and meta information which stipulates when and how primitive activities are used. Investigations into the composition of the proposed XSet and the sequences in which primitive activities are arranged in the XSet reveal that relatively best swarm performances arise when the proposed XSet triggers robotic devices to respectively, (a) drop specific levels of pheromone chemicals in specific quantities, (b) orientate towards relatively attractive directions based on the levels of pheromone around, (c) move along the direction of orientation, and (d) alternate between different internal states when it becomes necessary. Besides prescribing a new robotic device control paradigm, this work inspires further developments towards generalized robotic device control standards.