Optimization of replenishment strategy for a VMI-based apparel supply chain

Abstract

Vendor Managed Inventory (VMI) is a collaborative strategy between the retailers and manufacturers to optimize the availability of products through a continuous replenishment approach. It has received considerable attention in the apparel industry. Apparel manufacturers adopting VMI strategies make decisions in their management systems with information shared with the retailers. With the responsibility for managing replenishment given to the manufacturers, one challenge facing the apparel supply chain is to balance the benefits of different parties in the supply chain by minimizing the stock out level of the retailers while maximizing the production capacity of the manufacturers. For this purpose, a simulation-based optimization model for VMI replenishment strategy in the apparel supply chain was proposed. It considers both the Customer Service Level (CSL) of the retailers and the constraints of the manufacturers' production capacity. This simulation based-optimization model consists of two main parts, namely, the simulation model for simulating the supply chain and the optimization algorithms for searching the optimal VMI-based replenishment strategy. The simulation model was developed to study the relationship between the replenishment strategy (replenishment cycle, lead time and replenishment quantity) and the performance of the supply chain (CSL and Inventory Turnover) in a VMI-based apparel supply chain. Fuzzy set theory was integrated into the simulation model so as to represent the forecasting error due to the dynamic customers' demands. The influence of dynamic and uncertain customers' demands on the performance of apparel supply chain was identified and examined using the proposed simulation model. Based on the data collected from the industry, the simulation model could generate the replenishment strategy in terms of replenishment quantity in each replenishment cycle. The simulation procedure was validated using the industrial data. For the simulation-based optimization model developed for the VMI-based apparel supply chain, the replenishment strategy generated by the simulation model was further optimized by using genetic algorithm (GA). Experimental results indicated that the proposed optimization of the VMI strategy could maintain the retailers' targeted CSL while improving the Production Capacity Balance of the manufacturers significantly. Validation of the optimization model was undertaken by comparing the performance of the optimized model with that of the industrial practice. To understand thoroughly about the current practice of the apparel supply chain adopting VMI-based strategy, in-depth experiments were conducted using the proposed simulation-based optimization model. Full factorial experiments were conducted based on the industrial practice. In the discussion of experimental results, the implications on the VMI-based replenishment strategy were described. Suggestions on how to improve the performance of the whole supply chain (CSL, Inventory Turnover and Production Capacity Balance) were given to the manufacturers and retailers in the VMI-based apparel supply chain based on the implications.