On considering journey time variability and passengers' path choice: an empirical study using Oyster card data on London Undeground

In this paper, we present an empirical study with the use of Oyster card data on London Underground, aiming at interrogating specifically on the following two issues: for one thing, the measurement of the interaction between passengers’ travel demand and their journey time variability by times of a day and between days; for another, the potential of inferring and modelling passengers’ path choice behaviour on public transport relying on smartcard data. Based on the evidence presented from statistical analyses of the Oyster card data, Weighted Oyster Journey Time (Weighted-OJT) is proposed for representation of passengers’ journey time distribution, which accommodates variations of the journey time correlating with unsteady travel demand given different periods. And it is demonstrated to be potentially having the Generalised Extreme Value (GEV) distribution. On this basis, the feasibility of revealing passengers’ path choice through the Oyster data is further discussed and future work is also suggested.