Priorização de inbound em centro de distribuição: estudo de caso em uma empresa de bens de consumo não duráveis
Lopes, Karine Fasolin
MetadataShow full item record
Companies in the distribution sector of non-durable consumer goods in Brazil suffer with the sales' concentration at the end of the marketing period, i.e., when a large amount of deliveries to customers occur in the last days of the month. This phenomenon can directly affect the company's costs as well as the level of service for lack of available products. Quite common scenarios that happen in distribution centers (DC) in this period are vehicles waiting to unload and the lack of products for delivery orders – it can also happen that the missing products are queued on hold in the trucks, waiting to be received. Considering the operations in distribution centers during this period is focused on the outbound of products, just some docks are available for unload. Based on this, the key issue is to determine which trucks must enter first at the docks to unload considering the necessity of products for the shipping operations and also the costs involved in this choice. Hence, the main objective of this study is firstly to propose a Mathematical Programming model to assists in prioritizing incoming vehicles, i.e., to determine which trucks must unload only considering the lack of product. Also, evaluate an extension of this model to minimize the costs of prioritization of vehicles, considering the lack of products, extra hours and demurrage per vehicle not received. The model is based on problems of vehicles assignation such as the gate assignments in airports and ships allocation in ports. Nevertheless, taking into consideration a distribution center of a company on the non-durable Consumer Goods Sector. The model was solved using the commercial package GAMS/CPLEX and the results were compared with the plans practiced by the company obtaining an average improvement for minimizing the lack of products of 37.1 % and 31% of minimizing costs. Furthermore, in order to verify the model fitness, some random instances were also performed. The model results generated adequate solutions to the problem, showing that the model is a viable tool for the inbound prioritizing of products.