Integrated Consolidation Optimization in E-commerce Supply Chain Networks
Vikas Bawa, MEng Student, Dalhousie Industrial Engineering Department
Many Statistics Canada surveys have shown that Canadian e-commerce sales are showing substantial growth each year. In the current best practice of distribution known as myopic consolidation, e-vendors are requested to consolidate their orders and deliver them to the nearest e-client partnered 3PL location.
Recently, researchers have proposed an alternative method called integrated consolidation, where e-vendors send FTL shipments to a strategic 3PL hub from where on, the orders are delivered by LTL shipments to the e-consumers. This not only reduces the transportation cost for e-vendors but also promotes cheaper and faster delivery to e-consumers by dynamically selecting drop-off locations.
In this research, we extend a single-DC integrated consolidation MIP model to multiple DCs. The resulting problem is NP-hard, since it involves both bin-packing and two-echelon routing. Surrogate LTL costs are used for the second stage of routing, i.e., from 3PL locations to e-consumer locations. We performed several computational experiments on small, medium, and large networks using randomly sized data. The problems were run using both a powerful local server and the Acenet high performance network. The impact of a valid inequality proposed to speed up the problem is investigated.
The main conclusions of this work are that a powerful local server performed better than Acenet. However, exact solutions are severely limited by the size of the network and number of orders. The valid inequality worsens computational time when the number of orders increases.
MA 310, 5269 Morris Street