Geographical implications of online shopping on physical distribution networks

Online shopping offers a competitive and attractive alternative to the conventional retail service by allowing us to order almost any product from anywhere and at anytime. Delivery of such online shopping products to individual customer locations depends on the reliability of logistics service that looks over the physical distribution of commodities. As the growth of the online shopping business is expected to increase the demand for parcel delivery service, it is imperative for the logistics service providers to have good insights into the cost behavior of their distribution networks with different levels of demand and to plan ahead for the adaptation of their network to meet future demands.;Unfortunately, only few analytical studies exist on the identification of the interdependent relationship between online shopping and physical distribution; and these are primarily concerned with the configuration of cost efficient distribution networks under a constant demand. Instead, this study discusses possible adaptive measures to an increase in online shopping demand, and it considers such modifications at the operational, tactical, and strategic levels of decision makings in the physical distribution network.;As the nature of daily freight flow in a large distribution network is so dynamic and complex, conventional mathematic models have limitations in representing and capturing such an environment. The discrete-event simulation model developed in this study shows how the overall cost efficiency and the network performance measures would change under various scenarios of parcel flow changes. To model the adaptation of the distribution network to the increasing demand of online shopping, we test three possible decision-making scenarios: (1) priority assignment by destinations at the main hub terminal (operational), (2) introduction of sub-hub terminals (tactical), and (3) increase in the capacity of the main hub terminal (strategic). The results of the simulation are calibrated in a GIS environment to indicate any delay in the delivery of products as well as to confirm the overall reliability of the logistics service. The feasibility of the simulation is tested with the logistics service data from an existing major distributor in South Korea.;The simulation study suggests that assigning a higher priority to parcels heading to a distant destination improves the overall reliability of the delivery service, although its performance rapidly diminishes when there is an abrupt and large increase in the amount of parcel flow. Introducing sub-hubs on the network relieves congestions in the main hub, but with the additional cost of maintaining the sub-hubs. From the standpoint of the overall cost efficiency, increasing the processing capacity of the main hub terminal turns out to be the best option, provided that the return-on-investment is expected to be great enough. These findings from the simulation model suggest that the existing physical distribution network is more likely to evolve to a more centralized network structure with increased capacity of transshipment facilities with the anticipation of imminent increase in online shopping demand.;Key words: E-commerce, online shopping, logistics service, physical distribution networks, discrete event simulation, Geographic Information Systems...