Forecasting using artificial neural networks and aggregate production planning and dynamic model of inventory control for rib and single knit fabric

Abstract

PT. ZYX is a textile and garment industry. The important department of this company is the knitting department that knits yarn into cloth, which will be used for the garment production of garment. Companies often experience shortages of stock to conduct production activities to meet consumer needs. In addition, the fluctuate demand sometimes does not match the available resources. This condition lead to the increasing of inventory costs. Based on the existing problems.The aim of this research are forecasting demand with Artificial Neural Network (ANN) method, calculating the production aggregate, controlling the cost of inventory control procurement, and designing a production planning inventory control system. Forecasting results are the input data to calculate the aggregate total production and the total inventory cost. Mixed Strategy is the best solution method for PT.ZYX, with a total cost of Rp 2,435,878,117.86, -. The best aggregate method for PT. ZYX is a mixed strategy method, because it gives the smallest total production cost of Rp 2,355,466,975. The inventory control is classified as dynamic inventory can save cost for 19.27% in RIB 1x1 product; 36.67% in RIB 2X1 product; 2.01% at RIB 5X2 product; 60.52% for single knit product.