Abstract:
The primary objective of this study is to establish a method for measuring bus travel
time reliability maintaining threshold levels at both the origin and destination points.
The threshold level was defined to be equal to half of the scheduled headway. Then the
concept of perceived waiting time was introduced. Perceived waiting time was derived
by subtracting half of the scheduled headway from the actual headway. Departure
activities were subsequently categorized into three distinct types: Early Departure (ED)
when the perceived waiting time was less than half of the scheduled headway, On-Time
Departure (OD) when the perceived waiting time equalled half of the scheduled
headway, and Late Departure (LD) when perceived waiting time exceeded half of the
scheduled headway. Similarly, arrival activities were classified into Early Arrival (EA),
On-Time Arrival (OA), and Late Arrival (LA), employing the same criteria. Then
crosstabulation analysis was employed to explore the association between departure and
arrival activity levels relative to day type (weekdays and weekends) and time of day.
Timekeeper records on scheduled departure time, actual departure time from the origin
bus stop, scheduled arrival time and the actual arrival time at the destination bus stop
were collected for around 1200 bus trips. Chi-square test was employed to test the
association between the departure and the arrival activities. In application to a suburban
feeder bus route of 255 between Mount Lavinia to Kottawa, out of the many, a notable
finding was that during peak hours on weekdays, buses arrived at their destination late
despite early departures. This observation challenged conventional assumptions about
bus travel time reliability, underscoring the necessity for a nuanced assessment that
accounted for both ends of the journey. Recommendations can be amended to consider
the scheduling, timekeeping at the noted delay points and operation control actions to
provide reliable travel for the users.
