A Catalogue of Robotic Device Actions for Creating Swarm Intelligence Ontologies

11 pages•Published: November 24, 2022

Abstract

Efforts to synchronize mobile robotic devices into desired formation are a niche research area. Most tips for programming plausible swarm intelligence are discretely taken from mathematics, physics, biology, chemistry, or nature. However, integration of the different cues into useful swarm intelligence systems is challenging. The notion of swarm intelligence ontologies is compelling. It captures theories, rules, policies, and meta-information about the creation of practical swarm systems. Swarm intelligence ontologies can establish the relationships between different swarm modelling paradigms to bring about generality. Such generality requires us to review various forms of swarm intelligence systems seeking to understand simulated robotic device actions that give rise to emergent behaviour. Two arms of such robotic device actions are noted. One set of robotic device actions are non-interactive. These actions are mainly taken from mathematics, physics laws, or other elitist methods. On the other hand, there are robotic devices actions regarded as interactive. Commonly, interactive robotic devices are bio- inspired. This article sequentially discusses these two categories of robotic device actions towards building a catalogue of actions to inspire the creation of swarm intelligence ontologies. Notably, most non-interactive robotic devices have abilities to recall the paths they previously followed. In other cases, robotic devices can use some form of language to share directional cues. However, the bulky of the literature points to robotic devices that can follow chemical tips towards the targets. A very small chunk of non-interactive robotic devices can rely on geometry, calculus, forces, beacons, or landmarks to orientate. Some interactive robotic devices can explicitly message pass. Some use environment mediated interactions. The type of data shared between such robotic devices is usually overtly connoted, including stacks, vectors, chemicals, forces, landmarks, or beacons. That way, the key robotic devices actions at individual level are apparently associated with reading stacks, interpreting language verbs, detecting chemicals, self-localizing, motion planning, or finding directions. This specificity in the characteristics of robotic device actions is the basis for the design of the swarm intelligence ontologies we envision.

Keyphrases: control models, ontology, robotic devices, swarm intelligence

In: Hossana Twinomurinzi, Nkosikhona Msweli and Tendani Mawela (editors). Proceedings of NEMISA Summit and Colloquium 2022: The Future of Work and Digital Skills, vol 4, pages 31-41.

Links:	https://easychair.org/publications/paper/m8bp
	https://doi.org/10.29007/m6kk