RRTCV

RRT*-Enhanced Long-Horizon Path Planning for AUV Adaptive Sampling using a Cost Valley

Overview

This repository contains the code and data associated with the research paper:

RRT-Enhanced Long-Horizon Path Planning for AUV Adaptive Sampling using a Cost Valley*
Authors: Yaolin Ge, Jo Eidsvik, André Julius Hovd Olaisen
Affiliation: Department of Mathematical Sciences, Norwegian University of Science and Technology (NTNU), Trondheim, Norway

This study presents an adaptive sampling method using autonomous underwater vehicles (AUVs) for long-horizon path planning. It introduces a flexible cost valley concept combined with a non-myopic RRT* planner, optimizing oceanographic sampling while ensuring real-time computations on AUVs.

Key Features

• Adaptive Sampling with Multiple Objectives: Combines variance reduction and classification error minimization.
• Non-Myopic Path Planning: Uses an RRT* strategy instead of traditional greedy (myopic) methods.
• Cost Valley Concept: Integrates operational constraints and informative sampling criteria into a unified weighted cost function.
• Field Deployment: Successfully tested in a 2.5-hour AUV mission in the Trondheim fjord, Norway.

Repository Contents

• code/: Implementation of the RRT*-based long-horizon planner and cost valley calculations.
• data/: Sample datasets used for simulation and field trials.
• notebooks/: Jupyter notebooks for analyzing simulation results and visualizing cost fields.
• docs/: Additional documentation and references.

Installation

Clone the repository:

git clone https://github.com/MASCOT-NTNU/RRTCV.git
cd RRTCV

Install dependencies:

pip install -r requirements.txt

Citation

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

@article{Ge2025,
  author = {Yaolin Ge, Jo Eidsvik, André Julius Hovd Olaisen},
  title = {RRT*-Enhanced Long-Horizon Path Planning for AUV Adaptive Sampling using a Cost Valley},
  journal = {Knowledge-Based Systems},
  year = {2025},
  doi = {10.1016/j.knosys.2025.113261}
}

Contact

For questions or collaborations, please contact:

• Yaolin Ge: [email protected]
• Jo Eidsvik: [email protected]
• André Julius Hovd Olaisen: [email protected]

---

This work was supported by the Norwegian Research Council (RCN) through the MASCOT project (305445). We thank NTNU AURLab and SINTEF Ocean for their support and data access.