Initial performance analysis on underside iceberg profiling with autonomous underwater vehicle

Off the coast of Newfoundland and Labrador, Canada, ice management is frequently necessary. The operation monitors icebergs´ positions, and manages threats posed by icebergs in the vicinity of offshore installations. In this paper, we present a performance analysis of different iceberg underwater profiling strategies using Autonomous Underwater Vehicles (AUVs), such as underwater gliders. Two strategies for profiling an iceberg with AUVs are proposed. The performance and outcomes are initially analysed in a simulation environment based on MATLAB. The iceberg shape used in the simulation derived from real-world iceberg measurements is provided by the National Research Council Canada. The waypoints in simulations are generated autonomously with a sonar model using a modified 3D Ramer-Douglas-Peucker (RDP) algorithm. The modified RDP filter is programmed to extract the feature-significant data in the points cloud and radially expand them with a standoff distance. Simulation results of spiral and vertical-transit profiling mode are included and compared. The sonar model will be implemented with stochastic estimation, and the collision avoidance features will be developed in the vehicle dynamic section. We are integrating a Tritech Micron mechanical scanning sonar into a Slocum underwater hybrid glider for field verification of this modeling.