Forecasting extractions in a closed loop supply chain of spare parts: An industrial case study

In healthcare industry, companies like GEHC (General Electric Healthcare) buy back their products at the EOL (End of Life) phase and reuse the spare parts composing them. This process is referred to as spare parts harvesting. The harvested parts are included in the spare parts supply chain which presents specific characteristics like the availability of critical parts and the intermittent demand behavior. Add to that, the unpredictability of the parts' supply capacity from bought back systems is a challenge for healthcare companies. The focus of this paper is to provide an accurate forecasting method of the harvested parts supply capacity for GEHC. To achieve this objective, a comparative study is carried out between statistical forecasting models. Then, a forecasting process employing the most accurate models is provided using TSB-Croston, the 12-month moving average, the best ARIMA model chosen with the Box-Jenkins methodology, and an introduced business knowledge based model. In order to improve the designed method accuracy, the statistical models' forecast is adjusted using contextual information. An error measurement based on a modified MAPE error is introduced to evaluate the forecast. By means of the designed method, the monthly accuracy is improved by 9%.