Modeling swarm robotics with knowlang

Abstract

Swarm robotics has emerged as a paradigm whereby intelligent agents are considered to be autonomous entities that interact either cooperatively or non-cooperatively. The concept is biologically-inspired and offers many advantages compared with single-agent systems, such as: greater redundancy, reduced costs and risks, and the ability to distribute the overall work among swarm members, which may result in greater efficiency and performance. The distributed and local nature of these systems is the main factor in the high degree of parallelism displayed by their dynamics that often results in adaptation to changing environmental conditions and robustness to failure. This paper presents a formal approach to modeling self-adaptive behavior for swarm robotics. The approach relies on the KnowLang language, a formal language dedicated to knowledge representation for self-adaptive systems.