Modeling the Container Selection for Freight Transportation: Case Study of Iran

Significant advantages of intermodal and containerized transport have increased the global interest to this mode of transportation. This growing interest is reflected in the annual volume of container cargo growth. However, the container transport inside Iran does not have a proper place. Comparing the count of containers entering and leaving ports with the statistics obtained from railway and road maintenance organizations showed that more than 77% of the containerized imports have been stripped at ports and dispatched toward their ultimate destinations outside containers. These statistics also showed that more than 81% of the containerized exports have transported to ports by means other than containers. The main purpose of this study was to identify the most important variables affecting the selection of containerized freight transport and non-containerized freight transport options by applying decision tree models on the road freight movement and a set of variables describes the differences between these two options. The final model representing the selection of containerized transport was developed by the use of CHAID, QUEST, C5 and C$R decision tree algorithms. The results showed that the decision tree built via pruned C5 algorithm provides the best accuracy and most sensible list of important parameters. High-value and perishable commodities showed the greatest potential for containerized transport. The most important policy factors that could affect the tendency of cargo owners to use containerized transport are tariffs and the status of destination (whether it is a port). Policies that could encourage cargo owners to use intermodal transport include setting a lower tariff on container handling, reducing the cost of loading and unloading, increasing the port facilities supporting the containerized transport, adjusting customs, and development of dry ports.

Keywords:

Container , freight transport , decision tree , port , tariff