Abstract
The design of many transit networks is based on transfer stations. Transfers allow better service connectivity and the avoidance of deploying unaffordable door-to-door services among origin and destination points of low demand. However, these potential improvements in terms of cost and performance are achieved at the expenses of forcing users to incur a temporal disutility at transfer facilities. In this paper, we consider that the quality experienced by users at transfers mainly depends on the temporal disutility. It consists of the walking time between loading areas, the waiting time for bus arrivals at the loading area, and the variability of those times. Hence, a procedure to measure the quality at the transfer areas is proposed, integrating the three relevant components: the physical layout of the transfer area; synchronisation between connected routes; and discrepancies between their actual and scheduled service. A quadratic expression of the previous concepts is justified as a valid approach to measure the total quality of the transfer by transit operators with available data. The calibration of the parameters can be obtained considering the customer perception, the consensus of experts, and the partial consensus of experts with regularity conditions. The proposed model can monitor the transfer performance attributes that depend on service operations and allows attributing the loss of quality to each component involved in the process. The new methodology developed through the paper was applied to Barcelona’s New Bus Network, although it would be generally applicable in other modes of transport.