Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE credits
Transit reliability is one of the key factors in running a successful transit system from both
passengers’ and operators’ perspective. To improve the reliability of a transit service a
performance analysis is necessary. There are several service measures that can be applied to
evaluate the performance of a transit service, both in relation to service punctuality and
service regularity. Punctuality can be considered of higher importance on low frequency
lines and regularity on high frequency lines. Bunching is used to describe how vehicles
occupying the same bus route tend to bunch up and consequently the reliability decreases.
For improving reliability several holding control strategies can be applied such as schedulebased
holding, where early vehicles are held at time points, and headway
where vehicles are held to retrieve even headways between consecutive vehicles.
This thesis provides an overview of several different performance measures that can
be analyzed using
Automatic Vehicle Location data (AVL) and Automatic Passenger Counters
data (APC) collected from bus vehicles. As a case study, bus line 1 in Stockholm was
analyzed. The line is a high frequency, inner city bus line, where schedule based holding is
the current holding control strategy.
The performance analysis included an analysis of service regularity, service
punctuality, dwell times, passenger boarding/alighting and load, and run times. A linear
regression analysis was applied to evaluate the effects of passenger activity on the dwell
The results showed that the overall service performance decreased along the line for
both directions. Vehicle trajectories revealed increased bunching along the line. The drivers’
compliance to holding analysis showed that there was room for improvement. Overall, the
analysis showed that the current holding control strategy does not retrieve headway
regularity and that the schedule for vehicle run times was too tight and needs revision.
Furthermore, switching to headway
‐based holding was suggested.
2012. , 114 p.