Discrepancy Between Hyperpath and Actual Route Choices Based on Smart-Card Data in Shizuoka, Japan

Abstract

Understanding passengers’ route choices plays an important role in public transport planning. As an alternative to traditional web-based surveys, smart cards, an emerging technology for fare collection, can be useful to obtain massive amounts of information over a long period of time. This paper determines how smart-card data can help us understand passengers’ travel strategies, by identifying factors influencing route choice behaviour within the bus system of Shizuoka, Japan. We also examine the discrepancy between hyperpath and actual route choices based on smart-card data together with the choice principles proposed by Luo et al. [17] and arrival time-based route assignment. Origin–destination (OD) pairs are analysed with the goal to determine the most appropriate model for the transit assignment problem. We found a discrepancy between hyperpath and actual route choices, which was attributed to the erroneous assumption of random arrivals of bus services; in fact, bus operation in the Shizuoka area is timetable-based. Consequently, passengers are likely to follow the bus schedule. Route choice flexibility was displayed by regular commuters, who did not strictly adhere to a single bus route even though they usually travelled according to the same origin–destination pair. This supports the concept of “hyperpath travellers”. A variety of factors, such as perceived crowding and uncertainty of services (delays), might also affect choices when bus routes are overcrowded and another bus is due shortly. Our findings aim to assist transport planners towards predicting traffic demand more accurately, and therefore enhancing the provided public transport services and determining influential factors for commuters’ travel strategies.