فهرست نشریات علمی دسترسی آزاد ایران

The advancements in technology coupled with swift changes in modern lifestyles and consumer habits, have resulted in a marked rise in online shopping, that these days represents a considerable share of the global market. Responding to the increasing online demands of the present era, online-retailers are seeking a flexible and cost-effective supply chain which can be optimized by the use of cross-dock facilities. This study investigates the implementation of vehicle routing problem to schedule order delivery for an online-retailer by ensuring timely and precise order fulfillment to improve responsiveness and effectiveness. The applications of machine learning techniques as a cutting-edge technology have improved optimization processes in dynamic real-time conditions. The data used in this research pertains to an online retailer in large-size, in the first phase, by employing machine learning techniques a distance matrix of all nodes is obtained from their geographical longitude and latitude, which is then reduced by clustering algorithm (K-Means). Subsequently, the reduced distance matrix is used in a developed mixed-integer programming model scheduling supply and delivery through cross-dock and vehicle assignment, while taking into account the customers desired time frames, permissible delays, and acceptable service levels for different customer categories. Results of exact solution of this model have confirmed that by changing the key parameters of the model, it is possible to achieve outputs that lead to further profitability and success of an online-retailer.