<![CDATA[PT. OGC faces inaccuracies in diesel and jet fuel forecasting, leading to inventory shortages, higher logistics costs, and potential public dissatisfaction, which also impacts Indonesia's energy self-sufficiency and oil-gas trade deficit. This research proposes improvements to diesel and jet fuel planning management at PT. OGC using the Six Sigma DMAIC methodology to identify root causes of inventory issues and enhance planning accuracy. A mixed-methods approach was employed, combining primary and secondary data with SIPOC analysis, forecast error measurements (MAD, MAPE, DPMO, sigma level), correlation analysis, fishbone diagrams, Pareto charts, and process capability indices (Cp, Cpk). Simulations validated the proposed solutions. Current forecasting accuracy is below acceptable levels: diesel MAPE at 5.89% (exceeding the 5% SLA tolerance) and jet fuel MAPE at 11.26%. Cp and Cpk remained below 1.0. Root causes included weak forecasting methods, misaligned turnaround schedules with demand, and emergency shutdowns. Improvement initiatives reduced forecast errors: the ARIMA–XGBoost hybrid model lowered diesel MAPE to 4.21%, and the SARIMA–LSTM hybrid model reduced jet fuel MAPE to 4.00%. The proposed improvements effectively reduce forecasting errors, increase production stability, and decrease stock problems. Recommendations include adopting hybrid AI/ML forecasting models, aligning turnaround schedules, digitizing preventive maintenance, and implementing a monitoring dashboard with regular audits.]]>
