Swarm technology for search and rescue through multi-sensor multi-viewpoint target identification

摘要

In this paper, we present preliminary results from an undergraduate research project involving hardware and software development for a cooperative team of autonomous ground vehicles performing a search and rescue (SAR) operation. Key features of the proposed system are a novel localization method and cooperative target identification/situational evaluation using extremely low-cost hardware and simple user interfaces. The localization method relies on the use of technology from an optical computer mouse, which provides an innovative, low-computation solution for robot localization even on rough terrain. The proposed coordination scheme relies on low-bandwidth local communication, well suited to SAR environments, but also provides sufficient unit autonomy for the unstructured environment. Targets are located, identified and verified through cooperation and coordination of multiple viewpoints and data streams. The resulting system provides excellent flexibility, robustness and efficacy in a low-cost package.