Hierarchical Path Planning

Abstract

A robotic swarm consists of a system formed by multiple robots that usually work cooperatively to perform a certain task. One important research topic in this area is navigation. Traditional approaches use planners that exploit the individual configuration space of each agent, generally causing problems in group behavior, such as lack of consistency and traffic congestions. This paper aims to formulate a hierarchical abstraction that controls the swarm, defining a geometric formation of known properties that allows automated control of group consistency. Thus, it is possible to solve the path planning problem from the exploitation of its configuration space, which is independent of the number of robots. We perform simulated and real experiments in order to evaluate the proposed approach. Results show that the abstraction is effective to define the group formation as well as increase its coherence during navigation.

Publications (available only in Portuguese)

• Santos, V. G.; Chaimowicz, L., “Planejamento hierárquico para enxames robóticos baseado na exploração do espaço de configurações”, Revista Eletrônica de Iniciação Científica (REIC), 12(3), CTIC, 2012. PDF
• Santos, V. G.; Chaimowicz, L., “Uso de hierarquias no controle de enxames robóticos”, Anais do X Simpósio Brasileiro de Automação Inteligente (SBAI), 2011, pp.557,562. PDF