Skip to main content

Schedule Adherence

Introduction​

Schedule adherence monitoring is a critical function for public transit operations. It measures how well buses follow their published timetables, identifying delays, early departures, and missed trips. The Schedule Adherence system provides real-time analysis and historical reporting to help agencies enhance reliability, passenger satisfaction, and service delivery.

Key Benefits:

  • Improved punctuality through real-time tracking and alerts
  • Better planning with adherence analytics across routes and trips
  • Enhanced service delivery monitoring
  • Configurable thresholds per agency requirements

Core Functional Capability​

The Schedule Adherence module provides the following features:

  1. Real-time Monitoring

    • Tracks vehicle positions against scheduled trip times using GTFS-Realtime and static GTFS feeds.

  2. Adherence Metrics Calculation

    • Computes deviation from scheduled times at the start and end of trips.
    • Classifies trips as Early, On-Time, or Late based on a configurable window (default: ±10 minutes).

  3. Trip Completion Tracking

    • Monitors each vehicle's progression and calculates completion percentage for every scheduled trip.

  4. Historical Logging

    • Stores schedule adherence data in a local database for later analysis.

  5. Report Generation

    • Outputs structured JSON files for each evaluation cycle, suitable for integration into dashboards or analytics tools.

  6. Multi-Agency and Multi-Route Support

    • Handles multiple transit agencies, routes, and vehicle types with dynamic configurations.

System Design / Architecture​

The system is structured around a data processing pipeline that:

  • Fetches real-time position data
  • Matches it with schedule data
  • Applies time deviation logic
  • Outputs structured results to disk
flowchart TD A[GTFS-Realtime] --> B[Data Collection] A2[GTFS Static Schedule] --> B B --> C[Vehicle-Schedule Matching] C --> D[Adherence Calculation] D --> E[Output]

Algorithm​

The Schedule Adherence algorithm includes the following steps:

  1. Data Collection

    • Real-time data is fetched from GTFS-Realtime feeds (vehicle positions, timestamps, route IDs).
    • Schedule data is loaded from GTFS static files (trip start/end times, route mappings).

  2. Vehicle-Schedule Matching

    • Vehicles are matched to scheduled trips based on route IDs and closest timestamp alignment.
    • Only active trips are considered based on service calendar and time of day.

  3. Trip Start/End Interpolation

    • Since GTFS-RT does not provide explicit trip start or end markers, the system interpolates these using ETA data.
    • For terminals, we estimate the start and end time based on when a bus would have reached the first or last stop using its ETA.

  4. Adherence Calculation

    • Time deviations are calculated for:
      • Scheduled vs. actual trip start
      • Scheduled vs. actual trip end
    • Trips are labeled as:
      • Early: >10 minutes before scheduled
      • Late: >10 minutes after scheduled
      • On-Time: Within ±10 minutes window
    • These thresholds are configurable to suit each agency.

  5. Trip Completion Evaluation

    • Vehicle progress is calculated based on how many scheduled stops were reached.
    • Completion percentage is derived and logged.

  6. Output Generation

    • Adherence summaries are saved as structured JSON.
    • Each file contains trip-level metadata, timestamps, vehicle info, route ID, and adherence classification.

Input Data Requirements​

To ensure proper functioning of the logic, the following data inputs are expected:

  • GTFS-Realtime Feed:

    • vehicle_id
    • route_id
    • trip_id
    • timestamp
    • position (latitude, longitude)

  • GTFS Static Schedule:

    • trip_id
    • start_time, end_time
    • stop_sequence
    • route_id

  • Recommended Data Granularity:

    • GPS updates every 30 seconds or faster
    • Timestamps in UTC or timezone-aligned
    • ETA data accurate within 1-minute error margin

Output​

  • Output is written to a JSON file at each evaluation cycle.
  • Each file contains:
    • trip_id, vehicle_id, route_id
    • actual_start_time, scheduled_start_time
    • actual_end_time, scheduled_end_time
    • adherence_status ("early", "on-time", "late")
    • completion_percent

Open-Source Repo​

The system is available on GitLab:
🔗 gitlab.com/transport-stack/schedule-adherence

References​

Transit Standards