Comparative Analysis of Forecasting Methods on Water Production in PDAM XYZ

Abstract

Good production planning can meet the demand and availability of raw materials, as well as proper production planning and scheduling, production control, inventory control, and evaluation. PDAM XYZ produces clean water for the Banyumas Regency area. Water demand at PDAM XYZ is starting to exceed the production capacity limit, so a capacity addition plan is needed to meet the customer's clean water demand. The addition of capacity can be calculated by calculating demand forecasting. Thus, this study aims to determine the best forecasting method for making water demand forecasting calculations at PDAM XYZ and the estimated amount of water demand at PDAM XYZ in the next five years. The forecasting methods used are least square and regression. The accuracy values compared are MAD, MSE, RMSE, MAPE, and Tracking signal. The results of the comparison state that the regression method is better with a MAD value of 129938.4, MSE of 28536740000, RMSE of 168928.2, and MAPE of 0.05. So, in planning the addition of production capacity, the regression method can be used to forecast calculations as a reference for determining the additional production capacity. The forecasting results using the regression method show a value of 4,352,051 m3. Based on these results, it is expected that PDAM XYZ will be able to map the amount of clean water demand so that customer water needs can be met.