Exportation to the cloud of distributed robotic tasks implemented in ROS

Cloud computing is a paradigm shift in computation that has been gaining traction over the recent years, which is supported by the increasing ubiquity of a reliable wireless connection to the Internet. Cloud robotics, which aims at bringing this principle to the field of Mobile Robotics, allows robots accessing seemingly unlimited external computation, thus being able to free onboard computation power and perform more complex tasks or tasks that were not able to run otherwise. This paper describes the migration of two multi-robot tasks previously implemented and tested in ROS by our research group - multi-robot SLAM and multi-robot patrolling - to a cloud robotics-based implementation using the Rapyuta framework [12], with the aim of studying the tradeoff between robots' computation load decrease and bandwidth usage increase. With this purpose, both simulations and experiments with real robots were conducted.