Yapay Sinir Ağları ile Teslim Süresi Tahmini ve Savunma Sanayinde Uygulaması
Date
2023-02Author
Baltacı, Erdem
xmlui.dri2xhtml.METS-1.0.item-emb
Acik erisimxmlui.mirage2.itemSummaryView.MetaData
Show full item recordAbstract
The competition among the companies operating in the field of production is increasing day by day. Businesses constantly strive to improve themselves to have a say in the competitive market. Lead time is one of the most important performance indicators that businesses want to calculate with high accuracy and improve in order to meet customer expectations. The time elapsed from the receipt of the customer order to the delivery of the product is called the lead time. In this study, a model has been developed to predict the lead time of the product by using artificial neural networks (ANN), including the production time and raw material supply time. With the developed model, the lead time to be promised to the customer can be estimated with an acceptable accuracy. This will ensure on-time delivery and high customer satisfaction. The model has a structure that estimates the lead time by considering the raw material supply time, and in this respect, it is unique in the literature. It can be easily applied in production systems where production processes and product tree are complex. The developed model was run by setting up different scenarios with the data of a product produced in a company operating in the defense industry, and the delivery times and raw material supply times were analyzed. In order to improve the delivery time, it has been suggested to keep a safety stock for the raw materials that cause the lead time to be extended. It has been stated that by keeping the safety stock for the first 30 raw materials with the highest delay, an improvement of 70% can be achieved in the delay time and 77% in the penalty payment due to delay. In this respect, the prediction model is used not only as a decision-making tool, but also as an improvement tool. In this study, ANN applications have been explained in more detail than other studies in the literature. In this regard, this study is a guide that shows the steps to be followed in ANN estimation applications.