Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Design and Experimental Validation of an Autonomous USV for Sensor Fusion-Based Navigation in GNSS-Denied Environments

Samuel Cohen-Salmon, Itzik Klein

TL;DR

The paper tackles the challenge of evaluating sensor fusion-based navigation in GNSS-denied maritime environments by delivering MARVEL, a modular, low-cost autonomous USV built from off-the-shelf components and open-source software. It presents a comprehensive design and validation workflow, including a robust mechanical/electrical assembly, multi-sensor fusion capability (DVLs, EM logs, IMUs, MRU-P RTK-INS), and a software framework centered on ArduPilot and MAVLink for real-time experiments. The authors demonstrate field validation across calm and challenging seas, emphasizing safety, modularity, and open interfaces to support rapid experimentation and AI-driven navigation research. This platform enables researchers to test high-frequency data acquisition, real-time sensor fusion, and autonomous control in authentic maritime conditions, facilitating dataset generation and algorithm development beyond simulations.

Abstract

This paper presents the design, development, and experimental validation of MARVEL, an autonomous unmanned surface vehicle built for real-world testing of sensor fusion-based navigation algorithms in GNSS-denied environments. MARVEL was developed under strict constraints of cost-efficiency, portability, and seaworthiness, with the goal of creating a modular, accessible platform for high-frequency data acquisition and experimental learning. It integrates electromagnetic logs, Doppler velocity logs, inertial sensors, and real-time kinematic GNSS positioning. MARVEL enables real-time, in-situ validation of advanced navigation and AI-driven algorithms using redundant, synchronized sensors. Field experiments demonstrate the system's stability, maneuverability, and adaptability in challenging sea conditions. The platform offers a novel, scalable approach for researchers seeking affordable, open-ended tools to evaluate sensor fusion techniques under real-world maritime constraints.

Design and Experimental Validation of an Autonomous USV for Sensor Fusion-Based Navigation in GNSS-Denied Environments

TL;DR

The paper tackles the challenge of evaluating sensor fusion-based navigation in GNSS-denied maritime environments by delivering MARVEL, a modular, low-cost autonomous USV built from off-the-shelf components and open-source software. It presents a comprehensive design and validation workflow, including a robust mechanical/electrical assembly, multi-sensor fusion capability (DVLs, EM logs, IMUs, MRU-P RTK-INS), and a software framework centered on ArduPilot and MAVLink for real-time experiments. The authors demonstrate field validation across calm and challenging seas, emphasizing safety, modularity, and open interfaces to support rapid experimentation and AI-driven navigation research. This platform enables researchers to test high-frequency data acquisition, real-time sensor fusion, and autonomous control in authentic maritime conditions, facilitating dataset generation and algorithm development beyond simulations.

Abstract

This paper presents the design, development, and experimental validation of MARVEL, an autonomous unmanned surface vehicle built for real-world testing of sensor fusion-based navigation algorithms in GNSS-denied environments. MARVEL was developed under strict constraints of cost-efficiency, portability, and seaworthiness, with the goal of creating a modular, accessible platform for high-frequency data acquisition and experimental learning. It integrates electromagnetic logs, Doppler velocity logs, inertial sensors, and real-time kinematic GNSS positioning. MARVEL enables real-time, in-situ validation of advanced navigation and AI-driven algorithms using redundant, synchronized sensors. Field experiments demonstrate the system's stability, maneuverability, and adaptability in challenging sea conditions. The platform offers a novel, scalable approach for researchers seeking affordable, open-ended tools to evaluate sensor fusion techniques under real-world maritime constraints.

Paper Structure

This paper contains 17 sections, 7 figures.

Table of Contents

  1. Introduction
  2. Core Design Principles
  3. System Design and Architecture
  4. Structure
  5. Propulsion
  6. Power
  7. Control System
  8. Communication
  9. Sensors
  10. Software
  11. Safety
  12. Summary
  13. Assembly and Experiments
  14. Mechanical and Electrical Assembly
  15. Software Integration and Debugging
  16. ...and 2 more sections

Figures (7)

  • Figure 1: Core design principles guiding the development of MARVEL: accessibility, ease of use, low cost, seaworthiness, autonomy, and open-source compatibility.
  • Figure 2: Main structural frame of MARVEL following welding and paint application, left to dry before assembly.
  • Figure 3: System overview diagram showing power, data, and wireless communication lines between major components of the MARVEL USV.
  • Figure 4: Software workflow in the onboard computer.
  • Figure 5: Field deployment of MARVEL under strong wind conditions.
  • ...and 2 more figures