DART-Pi-Tracker

View on Render: https://dart-pi-tracker.onrender.com

Live GTFS transit overlay for:

• Dallas, TX (DART): http://www.dart.org/transitdata/latest/google_transit.zip
• Hattiesburg, MS (HCT): https://api.transloc.com/gtfs/hct.zip

The app downloads both feeds, unzips/parses GTFS, and displays routes + moving vehicle dots on a grayscale map.

Features

• Dallas/Hattiesburg city toggle (single-city view)
• Route overlays from GTFS shapes (with fallback geometry)
• Moving vehicles based on GTFS schedules
• Simulated fallback movement when no trips are active at the current time
• Route/vehicle color matching
• Vehicle-type filters (bus, train, other)
• Dallas train-line dropdown filter (per-line multi-select)
• Light and dark mode toggle (dark mode mutes line colors)
• Manual refresh endpoint for feed reload

Tech Stack

• Backend: Flask + Pandas + Requests
• Frontend: Leaflet + vanilla JS/CSS
• Deployment: Gunicorn (Render-ready)

Requirements

• Python 3.11+
• Internet access

Local Run

python -m venv .venv
.venv\Scripts\activate
pip install -r requirements.txt
python app.py

Open: http://127.0.0.1:5000

Data and Behavior

• Feeds are refreshed on startup.
• You can force refresh with POST /api/refresh.
• Vehicle positions are schedule-derived from static GTFS (not GTFS-RT).
• TRE is excluded from Dallas data.
• If there are no currently active trips, a service-window simulation is used so vehicles still animate.

API Endpoints

• GET /health
• GET /api/cities
• GET /api/routes?city=dallas|hattiesburg
• GET /api/vehicles?city=dallas|hattiesburg
• POST /api/refresh

Deploy on Render

This repo includes render.yaml.

1. Push this repo to GitHub.
2. In Render, create a new Web Service from the repo.
3. Render will use:
   • Build: pip install -r requirements.txt
   • Start: gunicorn app:app
4. Deploy.

The app binds to 0.0.0.0:$PORT for hosted environments.