Production planning and capacity control with demand forecasting using artificial neural network (Case study PT. Dynaplast) for industry 4.0

Abstract

PT. Dynaplast is a manufacturing company engaged in the production of plastic packaging. There are problems with fulfilling demands and overproduction. This research was conducted at the Blow 1 Department for Arrow, Morning fresh, ZWT, and BBF products using demand data from January 2017 to December 2019. Other than that, it takes some additional data to do disaggregated aggregate planning, rough-cut capacity planning, material requirement planning, and capacity requirement planning. Based on the analysis used, used forecasting methods SMA, DMA, WMA, SES, DES, Linear Regression, Cyclic, Quadratic, Decomposition, and ANN, the method chosen for PT. Dynaplast is an (ANN) with the smallest error value, disaggregate aggregate planning using mixed scheduling with a total cost of Rp 4.311.305.125. The results of rough-cut capacity planning using CPOF, BOLA, and RPA methods, all of these RCCP methods have sufficient resources. Based on lot sizing calculations using the LFL, EOQ, POQ, PPB, Silver Meal, AWW, and LUC techniques, the best results are the Algoritma Wagner Within (AWW) method optimal total cost for raw materials of Rp 10.633.664.526 and capacity requirement planning. There are deficiencies in the mixing machine, which can be resolved by adding overtime in January.