Bu çalışmada, Tedarikçi Yönetimli Envanter (VMI) yaklaşımının tedarik zinciri performansına sağladığı faydaları incelemek için bir benzetim modeli kurgulanmıştır. Benzetim modelinde kapasite sınırı olan bir üretici, bir distribütör, bir toptancı ve bir perakendeciden oluşan dört kademeli bir tedarik zinciri yapısı dikkate alınmıştır. Durağan ve durağan olmayan talep yapılarının dikkate alındığı benzetim modelinde iki farklı yapıda tedarik zinciri canlandırılmıştır. İlk yapı, geleneksel anlayışla yönetilen (TSS) tedarik zinciri yapısıdır. TSS yaklaşımında, tedarik zinciri üyeleri arasında herhangi bir bilgi paylaşımı olmamakta ve birbirleriyle sadece siparişler vererek iletişim kurmaktadırlar. İkinci yapı ise, VMI yaklaşımının uygulandığı tedarik zinciri yapısıdır. VMI yaklaşımı çerçevesinde, perakendecideki envanterin yönetiminden toptancı sorumludur. Benzetim modeli sonucunda elde edilen sonuçlar, VMI yaklaşımının, tedarik zinciri maliyetini önemli ölçüde düşürürken, müşteri gereksinimlerini daha yüksek oranda karşıladığını göstermektedir. Ayrıca elde edilen bulgular; üretim kapasitesinin ve müşteri talebinde gözlenen belirsizliklerin VMI yaklaşımından elde edilen fayda üzerinde anlamlı etkileri olduğunu göstermektedir. İlk olarak üretim kapasitesi dikkate alındığında, üretim kapasitesinin çok sınırlı olmasının VMI yaklaşımlarından elde edilen faydayı düşürdüğü görülmektedir. Müşteri talebinde gözlenen belirsizlikler dikkate alındığında ise, müşteri talebindeki belirsizliğin artmasının VMI yaklaşımından elde edilen faydayı önemli düzeylerde azalttığı görülmektedir. Sonuç olarak, VMI yaklaşımının üretim kapasitesinin yüksek olduğu ve müşteri talebinde gözlenen belirsizliklerin düşük olduğu durumlarda daha faydalı olduğu görülmektedir.

In this study, a comprehensive simulation model is built to explore the benefits of a supply chain initiative called “vendor-managed inventory (VMI)” where the vendors are authorized to manage inventories at retail locations. Under VMI, retailer shares demand information to the vendor who should then make stock level decisions for its own organization and the retailer. In this article, specifically, we are interested in finding answers to the following questions: Under what conditions will VMI more beneficial? How will lead times, demand uncertainty, and manufacturing capacity affect the benefits gained from VMI? From this investigation, we will better understand the required conditions for more successful VMI applications. For this purpose, we consider a four-echelon supply chain consisting of a capacitated manufacturer, a distributor, a wholesaler, and a retailer. The retailer realizes customer demands from gamma distribution. Stationary and non-stationary customer demand structures considered to explore the impacts of different levels of demand uncertainty on the benefits obtained from VMI. In order to evaluate VMI benefits, two stylized supply chain structures are considered. The first situation is a traditional model (TSS) where there is no information sharing and all members in the supply chain independently plan and operate the supply chain. The second situation is a supply chain model with vendor-managed inventory (VMI) where the wholesaler (as being the vendor for the retailer for the given supply chain) takes the full responsibility of managing the retailer’s inventory. Order-up to policy is used for ordering decisions for both systems.A full factorial design of experiment constructed to analyze the impacts of different factors on the benefits obtained from VMI. These factors are demand uncertainty, manufacturing capacity restrictions, and lead times. The levels of the each factor determined as follows:• Three levels for the demand uncertainty; demand structure with no seasonal swings, moderate degree of seasonal swings, and high degree of seasonal swings.• Manufacturing capacity restrictions measured by “capacity ratio” which corresponds to the ratio of total capacity to total demand. Thelevels for the manufacturing capacity restrictions are capacity ratio of 1.10, 1.30, and 1.50. • Two levels of lead-time; which is 1 and 4 periodsMoreover, total cost for entire supply chain, and customer service level are the dependent variables considered as performance metrics in experimental design. Customer service level is the percentage of customer demand satisfied through the available inventory of the retailer.Since there is more than one dependent variable in the experimental design, MANOVA is conducted to analyze the experimental simulation output. The results of the experimental simulation output indicate that when compared to traditional supply chain, VMI decreases supply chain cost substantially. Compared to the traditional settings, the reduction in the total supply chain cost varies from 6.5% to 43.3% with an average around 17.4%. When the service level is considered, we see that in all cases VMI reaches higher levels of service levels. For example, customer service level increased from 94.4% to 95.9% on the average. These results lead us to conclude that VMI is always beneficial to the supply chain performance. Moreover, the results of the MANOVA indicate that manufacturer’s capacity restrictions and customer demand uncertainty have significant impacts on the benefits obtained from VMI. Firstly, when manufacturing capacity is considered, the findings indicate that as the available manufacturing capacity is higher, the benefits obtained from VMI are also higher. Secondly, when the demand uncertainty is considered, the results indicate that as the uncertainty in customer demand increases, the performance of supply chain with VMI decreases substantially.In conclusion, through comprehensive simulation experiment and subsequent analysis of the simulation outputs, we made the following important findings: • VMI significantly improve the performance of the supply chain under all conditions.• The benefits gained from VMI significantly influenced by the manufacturing capacity and demand uncertainty. Therefore, VMI is more beneficial under the conditions with is high level of manufacturing capacity and low degree of demand uncertainty.