VAG Stats Tracker

Custom PGSql Export format.

Data from the API is published under the Creative Commons Attribution 4.0 Int. license
URL: https://opendata.vag.de/

Download latest database export: Download - 27 January 2025 until 19 July 2025

Try it out now
Install

Goal

Features

ToDo:

• Track delays of everything
  • Bus
  • U-Bahn
  • Tram
• Live Map of ALL services
• Generate Heatmaps of delays
• Geoline API - Get exact WGS84 coordinates for lines with rails
• Mutch faster Stops API
• Toplist

Things i´d like to do but got not enough time for:

• Easy way to report delays or other stuff (Not directly to the company)
• Smart warnings of potential delays
• Link common delay causes to the average delay they cause
• Track Elevator malfunctions and duration
• Smarter departure lookup endpoint
• Smarter connection lookup endpoint
• Something like Flightradar24s "People on my plane"
• Smart alexa alarmclock (Alexa, wake me up so i´ll be in scool by 8 o´clock)
• New API Methods to extend the VAG PULS API (If i figure out how to install PGSQL Plugins, lol)
• More...

Pictures:

Real Time Map

Vehicle + Accessibility info

For private vehicles this information isn't avaible and the VAG also doesn't know all the Vehicles (Laziness on their part)
Bus: (Operator) Displays AC Avaibility, Wheelchair Spots and Fule type (Diesel, Gasoline, Gas or Electro)

Tram: (Operator) and Door configuration (Wheelchair accessible or not)

Delay Heat Map

Displays all delays of a date or timespan on the map

API

Avaible at: /api/v1

Stops

/search: Look up a stop, even the once VAG dosn´t know because the data is broken
Parameters:

• Haltestellenname: String
• VGNKennung: Int
• VAGKennung: String
• Latitude: Int -Min/Max=180 optional
• Longitude: Int -Min/Max=180 optional
• Produkte: String (Bus,Tram,Ubahn)

/location:
Parameters:

• Latitude: Int -Min/Max=180 optional
• Longitude: Int -Min/Max=180 optional
• Radius: Number -Max=40000KM

Geolines

/:line Possible: 4,5,6,7,8,10,11,U1,U2,U3

Toplists

/delay/:list Possible: by_stops, by_lines, by_vehicles
Parameters:

• at: String (YYYY-MM-DD)
• from: String (YYYY-MM-DD) - Not fully added
• to: String (YYYY-MM-DD) - Not fully added
  /vehicle/distance/:at: Returns all traveld (air distance) by Fahrzeugnummer (So no UBahn since that refers to KursID in this case, PVU, or connections that return -1 in the PULS API Response)
• at: String (YYYY-MM-DD) (Betriebstag)

Live

/all Returns all current trips
/map Parameters:

• Linie: String

Heatmap

/:
Parameters:

• at: String (YYYY-MM-DD)
• from: String (YYYY-MM-DD) - Experimentell
• to: String (YYYY-MM-DD) - Experimentell

Statistics

/lines: Returns all autodetected lines
/delay/line: Array of avrage delays, split into 1h time buckets, for a line.
Parameters:

• line: String (Alphanumeric)
• days: Int (1-365)
  /delay/percentage/week/product: Avrage delay per week per product
  Parameters:
• outerNegBound: Int (-3600 - +3600)
• outerPosBound: Int (-3600 - +3600)
• lowerInnerBound: Int (-3600 - +3600)
• upperInnerBound: Int (-3600 - +3600)
• years: Int (1-10)
  /delay/percentage/month/product: Avrage delay per month per product
  Parameters:
• outerNegBound: Int (-3600 - +3600)
• outerPosBound: Int (-3600 - +3600)
• lowerInnerBound: Int (-3600 - +3600)
• upperInnerBound: Int (-3600 - +3600)
• years: Int (1-10)

Vehicle

/history/:at/:id
/servedStops/:at/:id
/servedLines/:at/:id
Parameters:

• id: Int (Fahrzeugnummer)
• at: String (YYYY-MM-DD) (Betriebstag)

Install

Windows

Requirements:

• Docker Desktop
• NodeJs 20
• PGAdmin (or other PG Interface)
• Some IDE
• Redis Insight (Optional)

Steps:

1. Create containers with docker compose up -d in some powershell window in the root of the repo.
2. Use SQL Tool (PGAdmin) to generate a DB. Default user admin and password adminpassword
3. You need to make a .env Can be fond below File in all programs you wanna use:
   • FahrtenProcessor (Tracks a ongoing trip)
   • FahrtenScanner (Finds new trips)
   • WebService (Provides /metrics endpoint and all API routes)
4. Run programs in order, start al programs from their root:
   • Go into /WebService and run node index.js wait till it its ready
   • Go into /FahrtenScanner and run node index.js (If you already had it running run node /tools/clearRedis.js to wipe the cache)
   • Go into /FahrtenProcessor and run node index.js
5. You know its working when Scanner shows a lot of things realy fast, and Processor starts picking up all the jobs over the next couple of seconds

EnvSetup

WebService

# Application
APPLICATION=WebService
DOMAIN=http://localhost
FALLBACKLANG=de

# Logging
LOG_LEVEL=4# 0 = error, 1 = warning, 2 = info, 3 = debug, 4 = system
LOG_TYPE=console# console or stdout
LOG_COLOR=true# true or false
LOG_TEMPLATE=# leave empty for default or enter a custom template
LOG_STACK=false
SENTRY_DSN=#Optional

# Database PG
DB_HOST="localhost"
DB_PORT=5432
DB_NAME="<Your Generated DB Name"
DB_USER="admin"
DB_PASSWORD="adminpassword"

# Database Redis (Cache)
REDIS_USER="default"
REDIS_PASSWORD="example"
REDIS_HOST="localhost"
REDIS_PORT=6379
REDIS_DB=0

# Webserver
GLOBALWAITTIME=0 # Delay in ms until /src is loaded
EXTRAERRORWEBDELAY=0
PORT=80 # Port for webserver

# Security
# HashSalts can be modifyed any time but will not apply to existing hashes... also keep in mind 0 will disable the hash, then all hashed IPs will not work because its switches to direct IP comparison
SALTROUNDS=12 # Salt rounds for password hashing
WEBTOKENLENGTH=64 # Length of the Web Token
WebTokenDurationH=96 # How long the Web Token is valid for in hours
Web2FAValidForMin=5 # How long the 2FA token is valid for in minutes

# Limiter
DECREASEPERMIN=120 # Used for Limiter

# Proxy Settings
# This can be exploited if your proxy does not overwrite the headers
# It will also display a warning if proxyed requests are comming in but are not enabled here
CLOUDFLARE_PROXY=false
ANY_PROXY=false

# Cache Settings
CACHEDRIVER=local # Can also be redis

# HyperExpress Settings (Webserver)
HE_FAST_BUFFERS=false # If true, will use allocUnsafe

FahrtenScanner & FahrtenProcessor

APPLICATION=FahrtenScanner #or FahrtenProcessor

# LOGGING
LOG_LEVEL=4
LOG_TYPE=console# console or stdout
LOG_COLOR=true# true or false
LOG_TEMPLATE=# leave empty for default or enter a custom template
SENTRY_DDSN=

# GENERAL
PRODUCTS=UBahn,Bus,Tram # Things you wanna track
WATCHDOG_TIMEOUT=60

# Database PG
DB_HOST="localhost"
DB_PORT=5432
DB_NAME="<Your Generated DB Name"
DB_USER="admin"
DB_PASSWORD="adminpassword"

# Database Redis (Cache)
REDIS_USER="default"
REDIS_PASSWORD="example"
REDIS_HOST="localhost"
REDIS_PORT=6379
REDIS_DB=0

SCANBEFORE=30 # Scan before x seconds
SCAN_INTERVAL=5 # Scan every x seconds
ERROR_EXPIRE=600 # Expire errors after X seconds

Project (Deliverd Nov 2024) - Just no longer got time to play with it :/

Unlike my other projects this project will be built around my learning curve to cloud like software architecture.
Target Hardware:

• 4 Nodes each with:
• 8x ARMv8 SoC
• 16 GB DDR4
• 1 TB NVME
• 1 Gbit/s LAN