Multiple regression models were developed to make separate estimates of the number of bus passengers embarking during the morning and those during the afternoon peak periods along specified bus-route segments in Metropolitan Manila. The dependent variable in the models was log-transferred to achieve normality of data. Since limited bus volume surveys were performed, the distribution of time spent by buses along a particular segment was used to describe the distribution of bus volume along it. To estimate the volume of embarking passengers along the particular bus-route segment, the average bus volume along the segment was multiplied by the onboard survey results of the number of embarking passengers per bus along the segment. Thirteen proposed new bus routes were identified using the Equilibre Multimodal, Multimodal Equilibrium (EMME/2) transport modeling software. The developed models were then applied to the proposed new bus routes to predict bus passenger embarking on a per-route segment basis. The models were able to satisfactorily determine the number of embarking passengers.
