Validating and calibrating a destination estimation algorithm for public transport smart card fare collection systems

Data from smart card fare collection systems has proven to be very useful to public transport planners. These systems provide a continuous flow of data on transactions made on networks; hence it helps to better understand customer (card) travel behavior, and the data can also be used to characterize and model general ridership, customer loyalty, and network performance indicators. But many systems only record the entrance ('tap-in') transaction in the system. There is a need to estimate the exit ('tap-out') location to have origin-destination trip information. In this paper, we use tap-in/tap-out smart card data from Brisbane, Australia, to calibrate and validate a trip destination estimation algorithm developed for Canadian data. Results show that the algorithm has an accuracy of 79% within an acceptable distance of 400 m. The proposed calibration method helped to solve 1.4% more destinations.