Cooperative Artificial Intelligence for underwater robotic swarm

Underwater Robots such as Autonomous Underwater Vehicles (AUVs) and Remotely Operated Vehicles (ROVs) has played an important role in many tasks, such as marine environmental monitoring, underwater resource exploration, oil and gas industries, hydrographic surveys, military missions, etc. Underwater robotic swarm is a team of cooperative underwater robots which focuses on controlling multiple underwater robots to work in an organic group. In contrast to a single underwater robot, underwater robotic swarm represents higher operation efficiency and better stability while executing complex tasks. However, it needs higher intelligence to realize complementary cooperation than a single robot. It is beneficial to researchers to present a comprehensive survey of the state of the art of cooperative research for underwater robotic swarm. We observe that the research of Artificial Intelligence (AI) for multiple underwater robots is still in an early stage. In this paper, we study different collaborative operation mode in detail, such as formation control, task allocation, path planning, obstacle avoidance, flocking control etc. We propose different classification frameworks for these research topics and it also can be used to compare different methods and help engineers choose suitable methods for various applications. To achieve better cooperative performance of underwater robots, there are several key factors, including multi-source heterogeneous sensing, cooperative communication and navigation, information fusion and decision. Moreover, cooperative AI for underwater robotic swarm has different kinds of interesting and helpful applications. Finally, several possible applied AI methods including meta-heuristic algorithms, deep learning method and distributed learning method are accomplishing to cooperation of underwater robotic swarm.(c) 2023 Elsevier B.V. All rights reserved.