PolarRoute

PolarRoute is a long-distance maritime polar route planning package, able to take into account complex and changing environmental conditions. It allows the construction of optimised routes through three main stages: discrete modelling of the environmental conditions using a non-uniform mesh, the construction of mesh-optimal paths, and physics informed path smoothing. In order to account for different vehicle properties we construct a series of data-driven functions that can be applied to the environmental mesh to determine the speed limitations and fuel requirements for a given vessel and mesh cell. The environmental modelling component of this functionality is provided by the MeshiPhi library.

Installation

PolarRoute is available from PyPI and the latest version can be installed by running:

pip install polar-route

Alternatively you can install PolarRoute by downloading the source code from GitHub:

git clone https://github.com/bas-amop/PolarRoute
cd PolarRoute
pip install -e .

Use of -e is optional, based on whether you want to be able to edit the installed copy of the package.

In order to run the test suite you will also need to include the test dependency group:

pip install --group test

NOTE: Some features of the PolarRoute package require GDAL to be installed. Please consult the documentation for further guidance.

Usage

PolarRoute operates by creating an environmental mesh, adding vessel performance characteristics, and optimizing routes between waypoints. Environmental meshes are created using MeshiPhi, which is installed automatically when installing PolarRoute.

Quick Start (CLI)

# Create environmental mesh
create_mesh examples/environment_config/grf_example.config.json -o mesh.json

# Add vessel performance model
add_vehicle examples/vessel_config/SDA.config.json mesh.json -o vessel_mesh.json

# Optimize routes
optimise_routes examples/route_config/traveltime.config.json vessel_mesh.json examples/waypoints_example.csv -o routes.json

Quick Start (Python API)

from meshiphi.mesh_generation.mesh_builder import MeshBuilder
from polar_route.vessel_performance.vessel_performance_modeller import VesselPerformanceModeller
from polar_route.route_planner.route_planner import RoutePlanner

# Create environmental mesh (using MeshiPhi)
mesh_builder = MeshBuilder(env_config)
mesh_json = mesh_builder.build_environmental_mesh().to_json()

# Add vessel performance to mesh
vpm = VesselPerformanceModeller(mesh_json, vessel_config)
vpm.model_accessibility()
vpm.model_performance()
vessel_mesh_json = vpm.to_json()

# Calculate routes
rp = RoutePlanner(vessel_mesh_json, route_config)
rp.compute_routes(waypoints_path)
routes_json = rp.to_json()

For more details, see the CLI documentation, examples, and the examples/ directory.

Required Data sources

PolarRoute has been built to work with a variety of open-source atmospheric and oceanographic data sources. For testing and demonstration purposes it is also possible to generate artificial Gaussian Random Field data.

A full list of supported data sources and their associated dataloaders is given in the 'Dataloader Overview' section of the MeshiPhi manual

Developers

Samuel Hall, Harrison Abbot, Ayat Fekry, George Coombs, David Wyld, Thomas Zwagerman, Jonathan Smith, Maria Fox, and James Byrne

License

This software is licensed under a MIT license, but request users cite our publication:

Jonathan D. Smith, Samuel Hall, George Coombs, Ayat Fekry, Michael A. S. Thorne, Derek Long, Maria Fox (2025) Path-Planning on a Spherical Surface with Disturbances and Exclusion Zones. Journal of Artificial Intelligence Research 82 (2025), https://doi.org/10.1613/jair.1.16746

For more information please see the attached LICENSE file.