Agent-based robot control design for multi-robot cooperation

Abstract

This paper presents an agent-based robot control (ARC) architecture. ARC-features a flexible realtime control system, which is suitable for multi-robot cooperative tasks. It also provides an efficient platform for building up a multi-robot system consisting of heterogeneous robots. In this paper, an experimental study of this architecture is investigated A cooperative exploration using two mobile robots will be demonstrated. In this experiment, one robot explores the environment by looking for a color-coded target and the other is responsible for task execution at the target position. While exploring in an unknown environment, the first robot, which is equipped with ultrasonic sensors for exploration, records its position as it sees deployed checkpoints. In a later phase, the second robot plans a path to the target directly using information passed from the first robot and get to the target position in an efficient way.