Task Allocation in Swarm Robotics pdfsubject

In this work, we propose a method for self-organized task partitioning and allocation. The current state of the art provides neither an abstract understanding of task allocation problems in self-organized swarm systems, nor tools for modeling and designing such systems. We present the first steps towards a unified method capable of exactly this. We analyse the problem and provide a structured approach to complex task allocation problems. The method relies on partitioning complex tasks in smaller, more manageable subtasks which can be tackled by a swarm of robots in a self-organized way. Two types of subtasks are identified and discussed. We present experiments to study the properties of these subtasks, and propose algorithms to achieve selforganized allocation to these subtasks as well as test the algorithms in practical, real-world applications.