Multi-robot mission optimisation : an online approach for optimised, long range inspection and sampling missions

Abstract

Mission execution optimisation is an essential aspect for the real world deployment of robotic systems. Execution optimisation can affect the outcome of a mission by allowing longer missions to be executed or by minimising the execution time of a mission. This work proposes methods for optimising inspection and sensing missions undertaken by a team of robots operating under communication and budget constraints. Regarding the inspection missions, it proposes the use of an information sharing architecture that is tolerant of communication errors combined with multirobot task allocation approaches that are inspired by the optimisation literature. Regarding the optimisation of sensing missions under budget constraints novel heuristic approaches are proposed that allow optimisation to be performed online. These methods are then combined to allow the online optimisation of long-range sensing missions performed by a team of robots communicating through a noisy channel and having budget constraints. All the proposed approaches have been evaluated using simulations and real-world robots. The gathered results are discussed in detail and show the benefits and the constraints of the proposed approaches, along with suggestions for further future directions.