Modeling the dynamic use and impact of travel information on activity-travel: (re)scheduling decisions

This paper describes a Bayesian framework for modelling the use and impact of travel information on activity-travel (re)scheduling decision choice under conditions of uncertainty. The approach uses an existing activity scheduling model (Aurora) as a (re) scheduling engine. Arentze and Timmermans' framework to model activity scheduling and rescheduling behavior under conditions of uncertainty is used to generalize Aurora to the case of uncertain, non-stationary environments. This framework is extended to model the choice of travel information simultaneously with (re)scheduling decisions. The proposed model takes into account the perception of credibility of information and learning based on experience. The paper outlines the framework and discusses ways to specify and test the models in future research.