DETERMINING DEMAND AND VENDOR SELECTION MODEL FOR FASHION PRODUCT IN 4TH PARTY LOGISTICS ENVIRONMENT

• Main Author: Permatasari, Pagitha
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/43516
• Institution: Institut Teknologi Bandung

The fashion industry is one of the creative industry sub-sectors that is developing rapidly and is quite influential on the Indonesian economy. Fashion products are quite complex products because one type of product can have many variations. The complexity of these fashion products is very risky for companies to have product waste that results in total costs. 4PL companies provide services for other companies to run their businesses from suppliers, manufacturing to warehousing. 4PL companies do not have their own resources and depend on other companies to meet the demand so 4PL companies must have a strong supply chain. Therefore, the task is how to choose vendors and suppliers also determine the allocation of the number of products that must be done in each selected vendor in order to meet demand in both quantity and quality. The selected vendors must be the vendors with the best performance. This research proposed a model to determine the number of demands and a mathematical model development of vendor and supplier selection in a 4PL environment by considering product quality. Decision variables in the form of number of upcoming demands, selected vendors, types the product allocated to the vendor, the number of products allocated to the vendor, the selected supplier, the type of material ordered from the supplier and the amount of material ordered to the supplier. The problem of determining the number of demands is solved by Artificial Neural Network (ANN) with the Feedforward-Backpropagation algorithm. Then vendor and supplier selection problem is done by Analytical Hierarchy Process (AHP) and mathematical model is done by Goal Programming weighting method with the aim of minimizing the deviation of each objective function. Sensitivity analysis is used to show the validity of the proposed model. This research offers global optimal result.