Exploring the Dynamics of Freight Delivery: A Comparative Analysis of Physical Internet and Conventional Supply Chain Methods

Abstract

This paper compares the effect of network configuration on the dynamics of freight delivery. A comparison is made between Physical Internet (PI) and conventional (CO) distribution for the line, tree, square, circle, and cluster topologies. A mixed-integer linear program (MILP) model is presented to optimize the trade-off between dispatching and inventory costs. Quicker dispatch reduces inventory costs but increases the cost of transportation. This model is then applied to both the PI and conventional distribution contexts for each topology to understand the conditions favorable to PI. The key factors affecting the performance of the two methods are the distance ratio and the capacity of vehicles relative to dispatch flows. In terms of topology, the line configuration is always beneficial to PI but in the case of the other topologies, PI is good when the distance ratios and the flow quantity in relation to vehicle capacity increase.