Background Electric vehicle (EV) adoption is growing rapidly, and cities need to decide where to place new public charge points so that they are accessible, fair and cost-effective. This project will focus on Cardiff as a case study and investigate how spatial data, road networks and optimisation techniques can be combined into a practical decision-support tool for planning on-street EV charging infrastructure.

Aim The aim of the project is to design and implement a GIS-based location-allocation system that helps planners explore different options for locating new on-street EV charge points in Cardiff. The system should highlight trade-offs between coverage, walking distance and equity across neighbourhoods.

Planned work

Data acquisition & preprocessing – Collect open spatial datasets for Cardiff (e.g. small-area boundaries and population, road network, existing public charge points, basic housing/parking characteristics). Clean and integrate these datasets in a common coordinate system, and derive a set of demand points (such as LSOA centroids or representative street segments) with associated demand weights.

Network & demand modelling – Build a routable road network graph and compute network distances between demand points and candidate charger locations. Define a simple EV demand index using population and housing/parking information.

Location-allocation modelling – Formulate one or more facility-location models (for example p-median to minimise average distance, or maximal-coverage to maximise the population within a given walking distance). Implement these models in Python using an optimisation library (e.g. OR-Tools / PuLP), running on top of the road network.

Prototype implementation – Develop a prototype decision-support tool with a spatial database layer (e.g. PostgreSQL/PostGIS or similar) and a simple web-based interface. The front-end will include an interactive map and basic charts to visualise demand, existing infrastructure, candidate sites and model results, and to allow “what-if” exploration by changing key parameters (number of new chargers, coverage radius, etc.).

Evaluation – Compare optimised siting scenarios with simple baselines (such as random siting or locating chargers only in the highest-demand areas) using quantitative metrics: mean/maximum distance to nearest charger, coverage rate and simple indicators of spatial equity between neighbourhoods. Summarise results with figures, tables and maps.