Introduction
On-time performance (OTP) is the most widely reported metric in public transit. It measures how reliably service arrives when scheduled — a direct indicator of passenger experience and contractual accountability. Yet agencies often struggle with inconsistent definitions, manual reporting processes, and results that vary depending on who calculates them.
This guide explains how to define, measure, and report OTP in a way that is consistent, defensible, and actionable.
Defining on-time performance
OTP is typically expressed as the percentage of observed arrivals (or departures) that fall within an agency-defined window around the scheduled time. Common windows include:
- Early/on-time/late — e.g., not more than 1 minute early and not more than 5 minutes late
- On-time only — e.g., within ±2 minutes of schedule
- Departure-based vs. arrival-based — measured at timepoint departure, timepoint arrival, or both
The definition must be documented and applied consistently across routes, reporting periods, and systems. Changing the window without noting it will make historical comparisons misleading.
Timepoints vs. all stops
Most agencies measure OTP at timepoints — selected stops where schedule adherence is critical — rather than at every stop along a route. Timepoints are usually major transfer points, terminal locations, or segment boundaries.
Measuring at every stop increases data volume and can produce different results, since minor schedule recovery between timepoints is normal. Analytics platforms should support agency-specific timepoint lists tied to GTFS stop IDs.
Data sources for OTP calculation
OTP can be calculated from several operational feeds:
- CAD/AVL — dispatch-grade arrival and departure events at stops
- GTFS-Realtime trip updates — predicted and actual times compared to static GTFS schedule
- Manual checks — field observations, increasingly replaced by automated methods
Automated calculation requires reliable mapping between vehicle events, trips, and scheduled stop times in GTFS. Misaligned stop IDs or trip IDs are the most common source of incorrect OTP figures.
Segmentation and reporting
Raw network-wide OTP rarely tells the full story. Effective reporting segments results by:
- Route and direction
- Time of day (AM peak, midday, PM peak, evening)
- Day type (weekday, Saturday, Sunday)
- Geography (division, garage, or corridor)
Segmentation reveals whether reliability problems are systemic or isolated to specific routes or periods — information operations teams need for targeted intervention.
OTP vs. other reliability metrics
OTP should not be interpreted in isolation:
- A route can show strong OTP while headways are irregular (bunching and gapping)
- High OTP with missed trips still fails passengers
- Excess wait time captures passenger impact on high-frequency routes better than schedule adherence alone
See our Transit KPI Guide for how OTP relates to headway adherence, service reliability, and other indicators.
Automating OTP reporting
Manual OTP reporting — exporting AVL data, joining spreadsheets, calculating percentages — is time-consuming and error-prone. Modern analytics platforms automate the workflow:
- Ingest GTFS schedule and operational feeds daily
- Apply agency-specific on-time windows and timepoint rules
- Calculate OTP by route, period, and day type
- Publish live and historical dashboards for operations and executives
Bus RT Insights automates on-time performance calculation with customizable rules, replacing periodic manual reports with continuous visibility.
Conclusion
Consistent OTP measurement depends on clear definitions, reliable data integration, and segmentation that supports operational decisions. Agencies that automate OTP reporting gain faster insight and more credible performance reporting.
Explore platform analytics or request a demo to see automated OTP dashboards.