This project proposes a lightweight, scalable method to locate optimal EV fast-charging stations using Voronoi geometry and a greedy algorithm. It maximizes coverage while avoiding overlap with current infrastructure.

The approach is inspired by Carlvo-Jurado et al. (2024), “Optimal location of electric vehicle charging stations using proximity diagrams”, published in Sustainable Cities and Society.
Read the paper

Data sources include OpenChargeMap, OpenStreetMap, and city population stats. The pipeline—built in Python with geopandas, shapely, and scipy—identifies high-potential points through spatial filtering and iterative selection.

Preliminary results show promising coverage improvements along strategic transport routes.

Presentation

This project focuses on classifying occupied properties (“okupadas”) using machine learning techniques. It involves incremental data scraping, cleaning, and feature engineering to enhance model performance. Geospatial features and area-specific occupation metrics were created to improve classification accuracy. With iterative improvements raising the F1-score from 37% to 59%.

Presentation

This is a test post to verify the layout and subtitle rendering.