How is the Reliability KPI calculated?

Shown below is the default behaviour if no customisation has been requested.

Reliability is calculated against scheduled timing points. In the example above, the journey would get a reliability rating of 75% as the vehicle was actually seen at 12 of the 16 timing points it was scheduled to be seen at.

Timing points are used to estimate the number of miles operated relative to those scheduled to be run

The Reliability KPI can be summed up as - how many scheduled timing points was the vehicle actually seen at?

What goes into the KPI Calculations?

KPI calculations are currently defined once for each customer, and can be broken into 3 different components:

Inputs - these cannot be customised, but include things like 1st Stop, Timing Points*, Last Stop*.
Thresholds - these are global values used across a customers network for early, late, & stop geofence sizes.
Weightings - these determine how much impact an “Input” has on your KPI result. (*Timing Points & Last Stop are combined into a single “input” at this time, but will be split in a future iteration)

To confirm what the current settings are for your KPI calculations, or to request a change please contact support.

A journey will be considered matched if one real-time heartbeat can be matched to a journey ID from the schedule data.