Predicting Chicken Egg Production Using the Seasonal Autoregressive Integrated Moving Average (SARIMA) Method at PT. PULAU MANDIRI JAYA FARM

Daniel Prasetyo

Journal of Engineering Science and Technology Management
Vol. 6 No. 1 (2026): Maret 2026

Chicken egg production at PT. Pulau Mandiri Jaya Farm for the 2021-2024 period showed significant monthly fluctuations (2,140-6,890 tons) due to seasonal patterns and feed management challenges, requiring accurate time series forecasting for optimal planning. This study aims to predict egg production using the SARIMA(1,1,1)(1,1,1,12) model to capture annual seasonality. This quantitative descriptive study analyzed monthly univariate data from 48 observations, with purposive sampling (80% training: 38 months; 20% testing: 10 months). Data analysis used Python statsmodels for stationarity testing, ACF/PACF identification, model fitting, and MSE evaluation. The results showed excellent model performance (MSE=0.3262; Ljung-Box p=0.44; Jarque-Bera p=0.65), accurately predicting 2025 production (3,886-6,873 tons). In conclusion, the SARIMA model enables precision feed stock planning (60-70% of production costs), improving operational efficiency and food security in South Sumatra.