Optimization of Cable-Road Layouts in Smart Forestry

Advanced multi-objective planning tools for designing cable-road corridors that balance productivity, ergonomics, and ecological impact in steep-terrain forestry. It provides a tool for simulating, optimizing, and indivudally adjusting cable corridor layouts.

Overview

This repository accompanies the dissertation LiDAR-based Multi-Objective Optimization for Cable Yarding in Steep Terrain. It combines high-resolution terrain data, geometric reasoning, and evolutionary optimization to automatically generate feasible cable-road networks. The toolkit supports researchers and practitioners who need to explore trade-offs between harvesting efficiency and environmental stewardship. See the figure below of an overview of how the components of the dissertation fit together:

Key Capabilities

• Generates feasible cable-road candidates through detailed geometric and mechanical checks.
• Optimizes corridor selections with single- and multi-objective approaches (AUGMECON, NSGA-II, manual weighting).
• Computes cost, ergonomic, and ecological indicators for each layout to quantify trade-offs.
• Provides interactive 2D/3D visualizations for inspecting cable corridors, anchor trees, and yarding distances.

Repository Layout

• src/main – Core Python modules for geometric computation, mechanical analysis, optimization routines, and plotting utilities.
• 01_Notebooks – Guided notebooks for preprocessing, optimization experiments, and interactive dashboards (compute_cable_corridors.ipynb, optimization.ipynb, interface.ipynb).
• 02_Figures – Figures, GIFs, and interactive assets used in publications and presentations.
• requirements.txt – Python dependencies for both notebooks and scripts.

Getting Started

1. Create and activate a virtual environment (Python ≥ 3.10 recommended).

   python -m venv .venv
   source .venv/bin/activate
   pip install --upgrade pip

2. Install project dependencies.

   pip install -r requirements.txt

3. Verify the installation by running unit tests or opening a notebook (see below).

Typical Workflow

1. Prepare data – The high-resolution forest datasets used in the dissertation are not bundled. Reach out if you need access for research use, or substitute your own GeoPackage/GeoJSON files following the structures in src/main/data_loading.py.
2. Generate cable-road candidates – Use src/main/cable_road_computation_main.py or the compute_cable_corridors.ipynb notebook to enumerate geometrically valid corridors and support trees.
3. Run optimization – Explore AUGMECON and NSGA-II Pareto fronts via src/main/optimization_execution.py or optimization.ipynb, adjusting objectives and constraints as needed.
4. Inspect results – Launch interface.ipynb to interactively filter Pareto-optimal corridors and visualize anchor configurations in 3D.

Notebooks for Exploration

• compute_cable_corridors.ipynb – End-to-end generation of cable road candidates with inline visual checks.
• optimization.ipynb – Multi-objective optimization workflows and Pareto front analysis.
• interface.ipynb – Plotly-based interactive selection and inspection of feasible corridor sets.
• debugging_and_visualization.ipynb – Utility plots for diagnosing mechanical checks and spatial filters.

Data & Reproducibility

The repository does not ship with proprietary forest maps or LiDAR-derived height models. To reproduce the dissertation results:

• Request access to the original datasets, or adapt your own spatial layers to match the expected schema (see src/main/classes_cable_road_computation.py).
• Update the data paths referenced in notebooks and scripts to point to your local copies.
• Re-run the notebooks to regenerate optimization results and figures.

Citation

If you build on this work, please cite the project using the metadata in CITATION.cff.

---

Questions, collaboration ideas, or data access requests are welcome—feel free to open an issue or reach out directly.