零件庫存預測與可維修零件補貨模式實證研究

Spare part buffer forecasting and repairable parts replenishment model: an empirical study

巫國興(Kuo-Hsing Wu)

指導教授 ： 黃惠民

中原大學/電機資訊學院/工業與系統工程研究所/博士(2017年)

https://doi.org/10.6840/cycu201700340

摘要

產品除了設計和功能外，售後服務是新產品導入市場成功或失敗的重要關鍵因素。零件可用度將影響服務的速度，也直接影響了服務品質。零件生命週期可以分為四個階段：準備、生產、生命週期終止 (EOL/End Of Life) 與服務終止 (EOS/End Of Service) 階段。在準備階段，由於缺乏歷史使用資訊，難以應用傳統統計技術進行需求預測。因此，需要採用不同的預測方法。新產品初始庫存的需求的問題可以通過延伸 Kahn (2002) 使用定性於預測的未來銷售的研究來完成。在本研究中，我們提出了一種用於預測零件需求準備階段的聯合定性方法 (JQM/Joint Qualitative Method) 方法。JQM的關鍵在於類比矩陣，及經由設定假設樂觀和悲觀的服務水平值分別為90％和60％。該標準化生成一種方法，使零件規劃人員可經由服務水準來取得零件需求預測值。
在零件生命週期中量產與EOL階段，本研究專注於可維修的零件，因為可維修零件通常屬於高單價。可維修的零件補充模型屬於為多階層庫存模型，可以是有限或無限多庫存的上層倉庫和下層倉庫的庫存管理模型。在實務，很多零件維修中心直覺認為服務點以固定週期退修故障零件的模式可便於維修工作的執行。然而，在實際上固定週期退修故障零件的模式在處理故障零件效率卻是不高。因此，本研究提出了一種稱為最短維修作業時間法 ( LTAT /Least Turn Around Time)，該方法利用最短維修作業時間作為退修作業底限。LTAT 是以零件故障率來預測可能的零件缺貨時間點，再以最短維修作業時間來反推其應退修時間點進行退修零件維修中心作業。使用此方法，零件維修中心修護品將即時到達服務點，補充已耗用量以避免任何缺貨。利用 LTAT 方法，與固定週期模型相比，總庫存需求和補貨週期將會減少。這表明 LTAT 模型可以提高零件庫存運行的效率和成本。

關鍵字

零件可用度 ； 零件初始庫存預測 ； 可維修零件補貨

並列摘要

Due to the increasing market competition, shorter product life, small profit margin and rising customer service demand, after service quality is getting more and more important. Besides product design and function, after service is a vital decisive role in the success or fail in the new product introduction. Spare part availability will affect the speed of service, which directly impacts the service quality. Spare part lifecycle can be classified into four stages: preparation, production, EOL (End of Life) and EOS (End of Service) stages. In the preparation stage, due to lack of historical usage information, it is hard to apply traditional statistic technique to perform demand forecasting. Therefore, a different approach to forecast usage is necessary. The problem to new product sales forecasting can be done by extending the research done by Kahn (2002) who uses qualitative forecasting. In this study, we propose a JQM (Joint Qualitative Method)) method for forecasting the preparation stage of spare part demand. The key in JQM is centered in the similarity matrix, and by assuming the optimistic and pessimistic service level value of 90% and 60 % respectively. This standardization generates a method which enable user to choose the service level for the demand condition.
For production and EOL stages, this study focusses in repairable spare part, since repairable spare part are likely the most expensive. Repairable spare part replenishing model can be classified into a multi-echelon inventory model with multi upper inventory hubs and lower level warehouses with limited or infinite stock. In current industrial practice, central component repair center intuitively assumes service point has a fixed cycle return for defects; this enables more convenient workload arrangement. However, a fixed cycle return model is not very efficient in handling the defect returns. Therefore, a LTAT (Least Turn Around Time) method is proposed in this study. The method leverages the minimum time in returning the repaired defect parts to the service point. The theory is based on the specific spare part failure rate to forecast the possible spare part shortage time; it is based on the minimum repair process time to reserve the time to ship defect parts to central component repair center. By this method, repaired parts should arrive at the service point in time and thus avoid any shortages. With LTAT method, the total inventory demand and replenishment cycle will be reduced compared to a fixed cycle model. This shows LTAT model can improve the efficiency of spare part inventory operation and the cost.

並列關鍵字

spare part availability ； initial buffer forecasting ； repairable spare part replenishment

參考文獻

Yang, K. H., and Wu, K. H., 2016. A spare part management model considering service level at a new product introduction: case of mobile phone, Journal of Quality (accepted).

Abdul-Jalbar, B., Segerstedt, A., Sicilia, J., and Nilsson, A., 2010. A new heuristic to solve the one-warehouse N-retailer problem, Computers & Operations Research, 37 (2), 265-272.

Agrawal, S., Singh, R. K., and Murtaza, Q., 2015. A literature review and perspectives in reverse logistics, Resources, Conservation and Recycling, 97, 76-92.

Armstrong J. S. (ed.),2009. Principles of Forecasting: A Handbook for Researchers and Practitioners, Kluwer Academic Publishers.

Ashayeri, J., Heuts, R., Jansen, A., and Szczerba, B., 1996. Inventory management of repairable service parts for personal computers: A case study, International Journal of Operations & Production Management, 16 (12), 74-97.