Data_Sheet_1_Improving Resource Management for Unattended Observation of the Marginal Ice Zone Using Autonomous Underwater Gliders.ZIP

• Main Authors: Zachary Duguid (9997634), Richard Camilli (1624681)
• Format: Dataset
• Language: unknown
• Published: 2021
• Subjects: Artificial Intelligence and Image Processing; Control Systems; Robotics and Automation; Adaptive Agents and Intelligent Robotics; Artificial Life; Computer Vision; Image Processing; Artificial Intelligence and Image Processing not elsewhere classified; Applied Ethics not elsewhere classified; Bioethics (human and animal); autonomous underwater glider; under-ice; long-range; onboard acoustic sensing; environment state estimation; marginal ice zone; adaptive control; energy efficiency; Arctic
• Online Access: https://doi.org/10.3389/frobt.2020.579256.s001

We present control policies for use with a modified autonomous underwater glider that are intended to enable remote launch/recovery and long-range unattended survey of the Arctic's marginal ice zone (MIZ). This region of the Arctic is poorly characterized but critical to the dynamics of ice advance and retreat. Due to the high cost of operating support vessels in the Arctic, the proposed glider architecture minimizes external infrastructure requirements for navigation and mission updates to brief and infrequent satellite updates on the order of once per day. This is possible through intelligent power management in combination with hybrid propulsion, adaptive velocity control, and dynamic depth band selection based on real-time environmental state estimation. We examine the energy savings, range improvements, decreased communication requirements, and temporal consistency that can be attained with the proposed glider architecture and control policies based on preliminary field data, and we discuss a future MIZ survey mission concept in the Arctic. Although the sensing and control policies presented here focus on under ice missions with an unattended underwater glider, they are hardware independent and are transferable to other robotic vehicle classes, including in aerial and space domains.