Quantifying Factors Influencing Urban Bus Passenger Boarding and Alighting Dynamics in an Emerging Economy

Abstract

A majority of the past studies on bus dwell time have been carried out in the context of developed countries, and the dwell time prediction models established in literature adopted the fixed-parameter assumption. The variations due to human factors such as passenger and driver behavior are not reflected under a fixed-parameter framework. This paper presents a methodology for rational estimation of passenger boarding and alighting time in the context of an emerging country using both fixed-parameter and random-parameter models. The methodology was demonstrated with a case study in the Kolkata metro area of India. The results established that the boarding and alighting time varies significantly across bus type, in-vehicle crowding, and size of the passenger group. In this context, random-parameter models were found to outperform fixed-parameter models in capturing the heterogeneity in bus and demand characteristics. The models were validated successfully with statistically acceptable errors.