Passenger arrival patterns and its implications for bus operation: The impact of schedule reading behavior on average waiting times at bus stops

Abstract

Inaccuracies in measuring passenger waiting times at bus stops can lead to significant inefficiencies in optimizing bus operation schemes. To address this issue, this paper introduces a refined methodology aimed at accurately representing passenger waiting times and estimating the distributions of passenger arrival patterns at bus stops, with a focus on low-frequency suburban buses, considering their schedule-reading behaviors. First, we conducted stated preference (SP) and revealed preference (RP) surveys to capture the factors affecting passengers’ arrival behaviors, revealing that bus vehicle headway and bus arrival punctuality (quantified as the standard deviation of arrival time deviations, SD-BATD) significantly influence passenger behavior. Second, we developed models to assess the proportions of schedule-reading passengers (SR-passengers) and their average waiting time (AWT) as well as standard deviation (SD-WT). By treating AWT and SD-WT as independent variables, we then characterized the arrival patterns of both SR-passengers and schedule-neglecting passengers (SN-passengers) using maximum extreme value and uniform distributions, respectively. Additionally, we conducted numerical experiments on bus headway optimization to validate the operational implications of the proposed model for bus services. The results demonstrate that the AWT model significantly reduces bus operation costs by up to 15.7% compared to the traditional assumption that AWT = 1/2 headway. This effect is particularly pronounced for routes characterized by lower demand and higher speeds, which are typical of low-frequency suburban buses. Furthermore, this paper highlights the importance of accurately estimating the passenger waiting times considering passenger schedule-reading behavior in optimizing bus services.