Home // ICAS 2020, The Sixteenth International Conference on Autonomic and Autonomous Systems // View article

A Human-on-the-Loop Autonomy Architecture for Resident-AUV Undersea Support Infrastructure

Authors:
Pedro Forero
LorRaine Duffy
Eric Hendricks

Keywords: Autonomy architecture; resident autonomous underwater vehicles; inspection, maintenance and repair; human-machine interaction.

Abstract:
The use of Resident Autonomous Underwater Vehicles (R-AUVs) is a necessary step towards increasing the safety and reliability of undersea infrastructure ranging from communication cables to oil pipelines and undersea observatories. Undersea Support Infrastructure (USI) for R-AUVs will provide docking, energy and communication services. Furthermore, it will be able to autonomously interact with R-AUVs, while enabling remote human operators to oversee, and in some cases direct, the R-AUVs and USI operations. This is particularly critical in situations where assured communications between operators and the USI are not guaranteed. This paper proposes an autonomy architecture for USIs that pursues a vertical and horizontal separation-of-concerns architecture-design approach and builds on well-documented autonomy and autonomic system design principles. Horizontal separation allows for configuration strategies and behavior policies to be defined, selected, executed and monitored by loosely coupled planning modules acting as arbitrators. Vertical separation enables decision making components to be grouped according to the timeliness of the decisions they must make. Our autonomy architecture features a dual, local and global, planning layer that provides an interface for the operator to interact with the USI, enables human-on-the-loop autonomy, and supports autonomous operations in situations where the communications between the operator and the USI are unreliable and intermittent. A use case for this architecture in the context of future at-sea inspection, maintenance, and repair services for the oil and gas industry is discussed.

Pages: 1 to 7

Copyright: Copyright (c) The Government of USA, 2020. Used by permission to IARIA.

Publication date: September 27, 2020

Published in: conference

ISSN: 2308-3913

ISBN: 978-1-61208-787-0

Location: Lisbon, Portugal

Dates: from September 27, 2020 to October 1, 2020