A proposal for a personalized trustable guidance system for urban railways considering passenger behavior attributes

Abstract

A “personalized guidance system” has been gathering attention as a service for railroad users that utilizes information and communication technology to guide passengers to the most appropriate train to board. In addition to commuters, there are also some private users who are willing to accept a delay in arrival at their destination if they can avoid congestion, and especially in recent years when infectious diseases are spreading, there are passengers who strongly desire to avoid congestion. The purpose of the individual guidance system proposed in this study is to provide individual guidance that meets the individual requirements of each passenger by classifying passengers into matrix groups according to their disutility for early and late arrival and for congestion, and by reflecting the behavioral characteristics of individual passengers in the guidance. Authors compared the results of two methods used by the system to assign trains to passengers: the “incremental assignment method,” in which passengers are assigned to trains one by one in a certain order, and the “user equilibrium method,” in which all passengers choose the best route for themselves. Authors also proposed an index to quantify passengers’ trust in individual guidance systems, and evaluated the passenger disutility and distrust caused by the guidance systems through an urban rail model. The results showed that when the order in which passengers are assigned to trains is random, both the overall disutility of passengers and their distrust of the guidance system are reduced, and that the disutility of passengers who are assigned to trains earlier is greater. This study confirmed that a guidance method that is easy to implement in society can reduce both overall disutility and individual trust to some extent, but it also revealed that passengers who make decisions early in the process can suffer greater disutility.