Shared autonomy for mapping and exploration of underwater environments using an HROV

Abstract

One of the main goals in robotics is to achieve full autonomy. However, for certain tasks, robots still lack the level of abstract reasoning that gives the ability to make decisions, plan ahead, and change actions during the execution of a process, these abilities are acquired at its best degree only by humans. In terms of unknown underwater environments, the combination of specialized robotics that can withstand harsh environments and the reasoning of a human operator can give effective results on exploration, inspection, and subsequent tasks. Standard solutions propose to use fully teleoperated Remotely Operated Vehicles (ROV)s or fully Autonomous Underwater Vehicles (AUV). Still, both solutions have their drawbacks: high operation costs and limitations due to the need of a physical connection in the case of ROVs, and limitations in the communications or problems while facing complex decision making in case of AUVs. These limitations can be overcome with the concept of shared autonomy applied to a Hybrid ROV (HROV). In this new paradigm, an operator leads the mission by selecting objectives at a high level, and a HROV executes them taking all the low level decisions. This paper explores the shared autonomy concept applied to an underwater exploration task, leading the first steps towards intervention missions.