The primary goal of transit agencies is to provide reliable, efficient, and productive transportation service. Transit reliability has been defined many times; for example in 1978, Abkowitz et al defined transit reliability as the availability and stability of transit service attributes and their effects on travel behaviour and on transit agencies’ performance travel behaviour and on transit agencies performance. Two years later, Turnquist and Blume described transit reliability as keeping transit vehicles on-schedule with uniform headways and consistent travel times. More recently, Kimple stated that transit reliability is a multidimensional phenomenon; consequently, there is not a single measure that can adequately address service quality. Bus transit reliability depends on several factors, including the route of travel, traffic conditions, time of day, and conditions at the bus stops along the route. The number of passengers alighting or boarding, fare payment method, dwell time, and the location of the bus stop also affect the overall reliability of bus transit service. Several research studies have been conducted on bus dwell time (DT), which is defined as the time interval between the opening and closing of the vehicle’s doors to serve passengers at a transit stop. This study defines a new variable: Total Bus Stop Time (TBST) which, in addition to DT, includes the time required for a bus to safely manoeuvre into a transit stop and the time consumed reentering the main traffic stream. It is thought that if the TBST is minimized, the overall reliability of bus transit along routes could be improved. It is clear that providing a reliable transit service is necessary in order to maintain an efficient and attractive system that increases users’ satisfaction and loyalty. Furthermore, reliable transit improves internal efficiency, reduces operating costs, and improves revenues by attracting and retaining users. Therefore, improving reliability is a benefit for both users and transit agencies, as it enables cities to achieve broader goals. This study was aimed at developing DT and TBST models for bus stops located at intersections and at midblock. The TBST models were developed using nonlinear optimisation methods. The study involved data collection at sixty bus stops, thirty of which were located at intersections, while the remaining bus stops were at midblock. The data was obtained for the morning, midday and evening peak hours during the period from January 2014 through September 2014. Data on the following variables were obtained at each bus stop: bus stop type, number of passengers alighting or boarding, DT, TBST, number of lanes on approach to the bus stop, presence of parking, and bus pad length. The data was analysed and all statistical inferences were conducted based on 95% confidence level. The results of the data analysed showed that, on average, both TBST and DT were higher at bus stops at intersections than at those located midblock. While the mean TBST was approximately 48 seconds at the bus stops at intersections, for the midblock bus stops, the mean DT and TBST were 21 and 35 seconds, respectively. The overall mean DT was determined to be 29 seconds. The regression models for the TBST and DT were determined to be statistically significant at the 95% confidence level based on the R2, F-Statistics, and model validation tests. The models could explain 67% to 96% of the variations in the data based on the R2 and adjusted R2 values. Tests including Kolmogorov-Smirnoff, normal probability, and residual plots were used to confirm the appropriateness of the models. The models were developed by bus stop type and by time of day. (Author/publisher)
Abstract