architecture.md

delay-prediction-service architecture

Note: How delay-prediction-service works is a detailed step-by-step walkthrough on how this service works.

This service needs access to the following infrastructure

• a PostgreSQL database, contains
  • the schedules data, generated from GTFS-Static data by scripts/build.sh
• a data directory, contains
  • trajectories, generated from GTFS-Static data by scripts/build.sh
  • information about the GTFS feed being used
• City of Herrenberg's Thingsboard instance
  • provides the realtime vehicle positions
• Stadnavi's MQTT broker
  • receives predicted GTFS-Realtime VehiclePositions & TripUpdates

delay-prediction-service processes input data (vehicle positions) as follows:

1. upon receiving vehicle positions from Thingsboard,
2. it will find current runs using the data in the PostgreSQL database,
3. match them with the trajectories stored in the data directory,
4. and send predictions to the MQTT broker.