Self-organizing Planner for Multiple Mobile Robot Exploration and Patrol

In this paper, we suggest a search and surveillance with mobile robots to collect information with minimum repeated coverage for efficiency. The problem applies to a variety of tasks, such as floor cleaning, map building, surveillance, security patrol operations, and search and rescue operations. Our method using the Reaction/Diffusion Equation on a Graph (RDEG) - based on-line strategy plans and re-plans according to incoming environmental information. The strategy can be applied to patrolling tasks after the complete exploration of an environment. The tasks are allocated among multiple mobile robots. The temporary leader plans a strategy for the other mobile robots in the team. The sensing and positional data of each robot are broadcast, and the environmental information is thus shared among the mobile robots. A simulation in various scenarios has been conducted using one to three mobile robots. The efficiency and effect of increasing the number of mobile robots in the team are discussed.