A Pedestrian Destination-Chain Choice Model from Bayesian Estimation of Pedestrian Activities using Sensors Data

Pedestrian modeling is emerging as a tool for designing new infrastructures and optimizing the use of current ones. Given sensor traces, we are interested in developing a dynamic model that can predict the destination chain of an individual in pedestrian facilities. As a first step, we developed a methodology to collect activity-episodes sequences from scarce data, directly modeling the imprecision in the measure. It generates several candidate lists of activity-episodes sequences associated with a corresponding likelihood.