Smooth path planning for autonomous pipeline inspections

This paper details a control method, used to guide an autonomous underwater vehicle (AUV) in the task of locating and tracking underwater pipeline and cables. The article details the simulation and experimental results of the navigation when searching for simulated pipes on the seabed in shallow waters. Sonar based pipeline detection and tracking starts by the acquisition of the sonar image and further realtime processing to distinguish the pipe from the seabed. Depending of the success of detection an AI based dynamic path planner must decide the next actions: a search or a track process. Pipeline searching involves a series of zig zag explorations of the area where the pipe is presumably located. Pipe tracking involves keeping the navigation at fixed distance from the pipe. The path following method proposed in this article is based on a combination of Lya- punov technique and PI controller acting over two horizontal thrusters dual torpedo AUV. The article details some of the stages involved in the detection of the pipe using sonar, the control method used for smooth path following, the main aspects of dynamic mission re-planning, and the results, both simulated and experimental for search process of a simulated pipe in shallow waters.