Travel Pattern Analysis Using Smart Card Data of Regular Users

Smart card fare collection systems are no longer a new trend: this will be the inevitable fare payment media in future transit networks. This paper addresses the use of smart card data to understand transit passenger behavior. Specifically, this paper (1) proposes a method of inferring boarding stops using Geographic Information Systems (GIS) and database management using Structured Query Language (SQL), and (2) analyzes the travel patterns of regular transit users overall and by fare type. The experiment of matching Automatic Fare Collection (AFC) data to Automatic Passenger Count with Vehicle Location (APC/VL) data clearly shows that intersection-level identification of passengers’ boarding stops can be successfully inferred. Through structured queries of the AFC data, regular users of a typical weekday are defined. The need for multi-day analysis to examine the travel and activity behavior of regular users is emphasized. Spatial-temporal characteristics then are analyzed by fare type and duration between transactions. In addition, the variability and regularity of transit use is investigated using data from the first transaction during each weekday.