A Novel Versatile Architecture for Autonomous Underwater Vehicle's Motion Planning and Task Assignment

TL;DR

This paper presents a comprehensive decision-making architecture for Autonomous Underwater Vehicles that integrates global route planning and local path optimization to enhance robustness, safety, and efficiency in dynamic underwater environments.

Contribution

It introduces a novel layered architecture combining global route planning and local path optimization with re-planning capabilities for AUVs in unknown environments.

Findings

Demonstrates improved mission productivity and safety in simulated underwater scenarios.

Shows robustness of the architecture against environmental dynamic changes.

Enhances AUV autonomy through integrated planning and re-planning modules.

Abstract

Expansion of today's underwater scenarios and missions necessitates the requestion for robust decision making of the Autonomous Underwater Vehicle (AUV); hence, design an efficient decision making framework is essential for maximizing the mission productivity in a restricted time. This paper focuses on developing a deliberative conflict-free-task assignment architecture encompassing a Global Route Planner (GRP) and a Local Path Planner (LPP) to provide consistent motion planning encountering both environmental dynamic changes and a priori knowledge of the terrain, so that the AUV is reactively guided to the target of interest in the context of an unknown underwater environment. The architecture involves three main modules: The GRP module at the top level deals with the task priority assignment, mission time management, and determination of a feasible route between start and destination point in a large scale environment. The LPP module at the lower level deals with safety considerations and generates collision-free optimal trajectory between each specific pair of waypoints listed in obtained global route. Re-planning module tends to promote robustness and reactive ability of the AUV with respect to the environmental changes. The experimental results for different simulated missions, demonstrate the inherent robustness and drastic efficiency of the proposed scheme in enhancement of the vehicles autonomy in terms of mission productivity, mission time management, and vehicle safety.