Impact of smartphone-delivered real-time multi-modal information

Abstract

En-trip mode switch decisions under the smartphone-delivered multi-modal information SMMI seem to have been rarely explored. This study investigated the impact on commute drivers' en-trip mode switch behaviour of SMMI. This was based on a stated preference SP survey in Shanghai which collected over 2000 observations of the choice between 'auto' and 'park-and-ride' P+R under SMMI. SMMI provides travel time for auto and P+R, delay for auto, cause of delay, P+R cost and comfort level of rail transit. A generalised estimating equations GEEs-based analysis was conducted to address the potential correlations between repeated observations from the same individual. Among the tested candidate GEEs models, the model with the 'unstructured' working correlation structure leads to the best fit for our SP data. Results showed that SMMI can significantly influence mode switch. Statistically significant explanatory variables in the model are gender, education level, usual commute mode, the number of sources of dynamic information of a driver access, frequency of driving, P+R use experience, seat availability in subway car, real time road condition, auto delay and P+R cost.