Production planning in the oil and gas industry is a key element for operational efficiency and response to changes in market demand. This research focuses on smart and adaptive strategies through the application of two main approaches: forecasting. The purpose of this study is to determine the most suitable forecasting model from the five forecasting models (Simple Exponential Smoothing, Naive Method, Simple Moving Average, Weighted Moving Average, and Exponential Moving Average) to be used in LPG production and calculate the value of forecasting in the next five periods. Using LPG historical data from 2017 to September 2023. Then the results are compared using forecasting error metrics such as MAPE and RMSE. It was concluded that the Simple Exponential Smoothing model showed a forecasting error value of 21.58% for MAPE and 72764.01 for RMSE. The Naive model has a forecast error value of 20.33% for MAPE and 78044.48 for RMSE. Meanwhile, the Simple Moving Average recorded a forecast error value of 20.28% for MAPE and 64449.76 for RMSE. On the other hand, the Weighted Moving Average shows a percentage error of 16.34% with an RMSE value of 48426.57. Finally, the Exponential Moving Average (EMA) shows an optimal level of accuracy, with a forecast error value of 16.01% for MAPE and 46046.42 for RMSE. Thus, from the five models evaluated, it can be concluded that the Exponential Moving Average (EMA) is the best model for forecasting LPG products, considering the lowest level of accuracy and percentage of forecasting errors. This study identifies the EMA as the best method in forecasting LPG production. The implication is a positive contribution to the accuracy of predictions and planning efficiency.