EV Charging Infrastructure WebGIS

A full-stack geospatial web application for visualizing public electric-vehicle charging infrastructure across Germany.

The project combines a responsive OpenLayers frontend with GeoServer, PostgreSQL/PostGIS and Docker. Charging-station data from the German Federal Network Agency is imported into PostGIS, published through OGC services and rendered as an interactive map with server-side filtering.

Overview

The application demonstrates a complete geospatial data pipeline:

Bundesnetzagentur CSV
        ↓
PostgreSQL + PostGIS
        ↓
GeoServer (WMS / WFS / GetFeatureInfo)
        ↓
OpenLayers web application

The prepared dataset contains more than 100,000 charging-station records with valid point geometries in EPSG:4326.

Features

Interactive map

• OpenStreetMap basemap
• GeoServer WMS charging-station layer
• AC and DC charging filters using CQL_FILTER
• Scale-dependent station rendering
• Station details through GetFeatureInfo
• Desktop hover tooltip
• Map reset and browser geolocation actions
• Live GeoServer connection status

Responsive interface

• Full-screen dark geospatial dashboard
• Desktop control panel
• Mobile filter drawer
• Accessible keyboard navigation and focus states
• Reduced-motion support
• Non-blocking status and error messages

Data and service layer

• PostgreSQL/PostGIS spatial storage
• GeoServer WMS and WFS publication
• Reproducible CSV import
• Automated GeoServer workspace, datastore, layer and style setup
• Safe and idempotent PowerShell setup scripts
• Validation for WMS, WFS, CQL filtering and GetFeatureInfo

Charging-station styling

The GeoServer layer uses a scale-dependent SLD style:

• AC stations: blue circular markers
• DC stations: orange circular markers with slightly larger symbols
• Marker sizes increase from regional to local map scales
• Semi-transparent fills and thin dark outlines preserve readability in dense areas

The style is stored at:

geoserver/styles/ev-charging-points.sld

The local setup process installs or updates the workspace style ev:ev-charging-points and assigns it to ev:ladepunkte automatically.

Technology stack

Geospatial and backend

• PostgreSQL
• PostGIS
• GeoServer
• OGC WMS
• OGC WFS
• CQL filters
• Docker Compose

Frontend

• OpenLayers
• Vanilla JavaScript
• Vite
• HTML5
• CSS3
• OpenStreetMap

Automation and data preparation

• PowerShell
• SQL
• Python
• CSV processing

Data source

The application uses public charging-infrastructure data from the Bundesnetzagentur Ladesäulenregister.

The repository includes the prepared file:

bnetza_min.csv

The larger raw export is intentionally not included. prepare_bnetza.py documents the preprocessing workflow used to create the reduced dataset.

Local setup on Windows

Requirements

• Docker Desktop
• Docker Compose
• A current Node.js LTS release
• npm
• Windows PowerShell or PowerShell 7

1. Start the services

From the repository root:

docker compose up -d

2. Configure PostGIS and GeoServer

.\scripts\setup-local.ps1

The setup script:

1. verifies that Docker is available,
2. waits for PostGIS and GeoServer,
3. creates the PostGIS schema when required,
4. imports bnetza_min.csv only when no valid rows exist,
5. configures the GeoServer workspace and datastore,
6. publishes ev:ladepunkte,
7. installs the custom SLD style,
8. validates WMS, WFS, CQL and GetFeatureInfo.

The script is safe to run repeatedly and does not duplicate imported data or GeoServer resources.

To intentionally rebuild the application table from the prepared CSV:

.\scripts\setup-local.ps1 -ResetData

This option affects only the application table. It does not remove the database or Docker volumes.

3. Start the frontend

cd web
npm install
npm run dev

Open:

http://localhost:5173

GeoServer is available at:

http://127.0.0.1:8080/geoserver

Important environment notes

Docker volumes

The postgis_data and geoserver_data named volumes are machine-local. Copying the repository to another computer does not copy the database contents or GeoServer configuration.

Running scripts/setup-local.ps1 restores the required local resources from the files stored in the repository.

Container networking

GeoServer connects to PostgreSQL through the Docker service name:

postgis

Using localhost or 127.0.0.1 inside the GeoServer container would point back to the GeoServer container itself.

Vite proxy

The frontend forwards /geoserver requests to:

http://127.0.0.1:8080

The IPv4 address is used deliberately because localhost may resolve to IPv6 ::1 on Windows.

Filtering behavior

The frontend applies server-side CQL filters to the WMS layer:

AC only:      anschlussart='AC'
DC only:      anschlussart='DC'
AC and DC:    anschlussart='AC' OR anschlussart='DC'
None:         1=0

For performance, AC stations are hidden below zoom level 9. The user's AC preference is preserved and restored when returning to a supported zoom level.

Project structure

ev-charging-map/
├── docker-compose.yml
├── bnetza_min.csv
├── prepare_bnetza.py
├── geoserver/
│   └── styles/
│       └── ev-charging-points.sld
├── scripts/
│   ├── setup-local.ps1
│   ├── import-bnetza.ps1
│   ├── configure-geoserver.ps1
│   ├── configure-geoserver-style.ps1
│   ├── validate-local.ps1
│   └── sql/
│       ├── create-ladepunkte.sql
│       └── import-bnetza.sql
├── web/
│   ├── public/
│   ├── src/
│   ├── index.html
│   ├── package.json
│   ├── package-lock.json
│   └── vite.config.js
├── .gitignore
└── README.md

Diagnostics

Check service status:

docker compose ps

Inspect recent logs:

docker compose logs --tail=100 postgis
docker compose logs --tail=100 geoserver

Build the frontend:

cd web
npm run build

Local development credentials

The credentials in docker-compose.yml are intended only for local demonstration and development. They must not be reused in a production environment.

Project context

This application was developed as an academic geospatial-processing project and later extended into a reproducible, portfolio-ready WebGIS system.

It demonstrates:

• spatial database design,
• geospatial service publication,
• OGC web standards,
• responsive frontend engineering,
• automated environment configuration.

Author

Developed by Ahmet Kislali.

License

No explicit open-source license has been assigned.

The repository is published for portfolio and educational presentation purposes.