Reproducibility Package for Drone-Based Medical Delivery Optimisation: Data, Models, and Visualisation Framework

OverviewThis repository provides the full reproducibility package for a research study on drone-based medical delivery optimisation. It includes datasets, simulation outputs, routing algorithms, and visualisation code used to evaluate system performance under multiple operational scenarios. The work investigates trade-offs between routing efficiency, mission success, infrastructure constraints, and demand variability, with a particular focus on comparing classical routing methods (such as Dijkstra and A*) against machine learning baselines and a proposed generative approach. ContentsThe repository is organised into the following components: Data: Raw and processed datasets used for simulation and evaluation Code: Python scripts and notebooks for data processing, modelling, and visualisation Models: Trained models and configuration files used in the experiments ReproducibilityAll results presented in the associated manuscript can be reproduced using the provided code and data. Each figure is generated through dedicated scripts or notebooks, enabling full traceability from raw inputs to final visual outputs. Key Features Comparative evaluation of multiple routing strategies under realistic constraints Robustness analysis across infrastructure scaling, demand surges, and low-data regimes Multi-objective frontier analysis capturing trade-offs between delay, success rate, and system capability Real versus synthetic data distribution comparisons for validation of generative modelling approaches Intended UseThis package is intended for researchers and practitioners working in: drone logistics and medical delivery systems optimisation and routing algorithms simulation-based evaluation of infrastructure systems machine learning for operational decision-making