Parts remanufacturing in the oilfield industry

Purpose – Provide a new technique for forecasting parts usage in remanufacturing operations and describe its application to oil field equipment operations. Design/methodology/approach – New “correlated forecasting preprocessing” equations were derived for extracting additional information from a matrix of historical parts usage data. They were applied to both synthetic data and actual data from a large oil field remanufacturer. Findings – The new equations were effective in extracting the necessary pre-forecasting data from both synthetic and actual data sets. The key to effective preprocessing is using the correlation information from the entire parts usage matrix rather than just rely on bill of materials relationships. Practical implications – Accurate forecasting is vital to manufacturing at all levels of aggregation from the smallest part to the entire facility and for all planning horizons from days to years. Forecasting is an operations manager's first line of defense in inventory control. The addition of a preprocessing step makes the application of traditional forecasting methods yield significantly better results. Originality/value – The concept of using the partial correlations present in a parts usage matrix, instead of attempting to apply bill of material relationships, is new. The resulting preprocessing equations are new. The value of improved forecasting is that it directly impacts manufacturing profitability.