The Farside Story

Abstract

Determining the proper location of bus stops is an important planning decision in the transit planning field. While previous efforts in the literature have suggested several advantages and disadvantages regarding certain bus stop placements, there has been little effort towards understanding the impacts of bus stop location on the transit system performance at the stop level of analysis. This article aims to evaluate the impact of bus stop location on bus stop time and stop time variation. It uses stop-level data collected from the Société de transport de Montréal (STM)'s automatic vehicle location (AVL) and automatic passenger count (APC) systems, in Montréal, Canada. The study findings show that stop times occurring on the near side of intersections are in average 4.2 to 5.0 seconds slower than stop times occurring on the far side of intersections, with no impact on stop time variation. A validation model is utilized to confirm the impacts of bus stop placements on stop time, using data from TriMet's automated bus dispatch system, in Portland, Oregon. This study offers transit planners and policy makers a better understanding of the effects of bus stop location on stop time and its variation in order to improve service quality while minimizing service variation at the stop level.

Identifying the Bias: Evaluating Effectiveness of Automatic Data Collection Methods in Estimating Details of Bus Dwell Time

“…Furthermore, the positive coefficient of the passenger friction factor shows the additional dwell time added on overcrowded buses with standees. These coefficients are consistent with findings from previous research (3,4,23).…”

“…For the remaining control variables in Models 1 through 3, dwell times at signalized intersections are slower compared to midblock stops or nonsignalized stops, which is consistent with previous studies (23). Dwells taking place at a bus stop with a connection to the metro are generally faster compared to a non-metro stop.…”

“…On the basis of previous research (3,4,23,28), a typical dwell time model is defined as follows: dwell time (s) = f (boardings, alightings, total passenger activity 2 , passenger load, friction, direction, time of day, lift usage, stop location, weather condition, fare payment method).…”

“…Total boardings are the number of people who boarded the bus, while passenger alightings are the number of people who exited the bus. The squared term of passenger activity is included to capture the potential nonlinear relationship of passenger activity, suggesting that each additional passenger requires less time to board than the passenger ahead (3,4,23). The passenger friction factor is used to capture the effect of an overcrowded bus with standees on dwell time and in this study was calculated as the sum of passenger activity and number of standees.…”

“…More recently, Diab and El-Geneidy evaluated the impact of bus stop location on dwell time and found longer dwell times associated with bus stops located on the nearside of an intersection (23). Furthermore, the authors estimated that passenger boarding added 3.3 s and each passenger alighting added 1.9 s to the total dwell time.…”