MOOS-IvP Code for Master's Research

Overview

This repository contains MOOS-IvP applications developed for master's research on COLREGs-compliant collision avoidance algorithms for Unmanned Surface Vessels (USVs).

Submitted Paper

"Toward Safe Autonomy at Sea: Implementation and Field Validation of COLREGs-Compliant Collision-Avoidance for Unmanned Surface Vessels"

• Authors: Douglas Silva de Lima, Gustavo Alencar Bisinotto, Eduardo Aoun Tannuri
• Submitted to: Journal of Marine Science and Engineering (October 2025)
• Institution: Mechatronics Engineering Department, Universidade de São Paulo

Modules

pContactSpawn

Spawns virtual contacts that approach the vessel in a simulated manner based on specified parameters.

• Contacts are generated to simulate COLREGs (COLREGS/RIPEAM) rules for testing purposes.
• Enables controlled encounter geometries (head-on, crossing, overtaking).

pAstarColAvd

Performs collision avoidance using a modified A* path-planning algorithm while respecting COLREGs rules.

• Implements COLREGs-aware forbidden sectors
• Features dynamic re-planning capabilities

pColAvdvo

Performs collision avoidance using a modified Velocity Obstacles (VO) algorithm with COLREGs compliance.

• Computes velocity-based collision cones in real-time
• Applies COLREGs-compliant maneuver selection

Spawning Contacts

Use the following command to spawn a virtual contact:

uPokeDB SPAWN_CONTACT="heading=270,relative_bearing=0,distance=500,speed=5"

Parameters:

• heading: Contact's heading in degrees (0-360)
• relative_bearing: Bearing relative to ownship (0-360)
• distance: Initial distance to contact in meters
• speed: Contact speed in m/s

Running Missions

Three mission bundles are provided, each testing a different collision avoidance algorithm:

1. col_avd_moos → Standard MOOS-IvP COLREGs Behavior (BHV_AvdColregsV22)
2. astar → Modified A* path-planning algorithm
3. velocity_obstacle → Modified Velocity Obstacles (VO) algorithm

How to Run:

cd missions/<mission_name>
./launch.sh

Key Features

• COLREGs Compliance: All algorithms implement International Regulations for Preventing Collisions at Sea (Rules 13-17)
• Field Validated: Tested in real-world conditions at Guanabara Bay, Rio de Janeiro, Brazil
• Three Encounter Types: Head-on, crossing, and overtaking scenarios
• Performance Metrics: CPA (Closest Point of Approach), distance traveled, mission time, and trajectory deviation

Experimental Results Summary

Algorithm	Mean CPA (m)	Mean Distance (m)	Mean Time (s)	Distance Increase (%)
COLREGs Behavior	30.0	179.4	174.7	27.2
Velocity Obstacle	13.0	193.1	186.0	37.6
A* Algorithm	30.0	203.5	202.3	46.5

System Requirements

• MOOS-IvP (tested with release 19.8+)
• Ubuntu 20.04 or later
• ROS2 (for hardware integration)

Citation

If you use this code in your research, please cite our work:

@article{delima2025colregs,
  title={Toward Safe Autonomy at Sea: Implementation and Field Validation of COLREGs-Compliant Collision-Avoidance for Unmanned Surface Vessels},
  author={de Lima, Douglas Silva and Bisinotto, Gustavo Alencar and Tannuri, Eduardo Aoun},
  journal={Submitted to Journal of Marine Science and Engineering},
  year={2025}
}

Acknowledgments

This work was partially supported by:

• FINEP (Brazilian Innovation Agency)
• CASNAV (Naval Systems Analysis Center)
• CNPq – Brazilian National Council for Scientific and Technological Development (process #300884/2025-7)
• TPN-USP Ship Maneuvering Simulation Center

License

This project is open-source and available under the terms specified in the repository.

Contact

For questions or collaborations:

• Douglas Silva de Lima: douglasli1995@usp.br
• Gustavo Alencar Bisinotto: gustavo.bisinotto@usp.br
• Eduardo Aoun Tannuri: eduat@usp.br

---

Repository: https://github.com/d2snc/moos-ivp-masters