Accessing End-Of-Supply Risk of Spare Parts Using Big Data
How to access the end-of-supply risk of spare pares using big data analytics
- Xishu Li
In sectors like aerospace, shipping, and defense, manufacturers and customers are focused on sustaining their products for prolonged periods. This attitude is due to the high costs and long lead times associated with new product development. As a result, the lifecycle of systems in these sectors often spans over 20, 30 or even more than 40 years. One of the main problems that these systems face during their lifetime is that parts of their system components are not supplied anymore. The procurement life of components, especially of electronic parts, is usually significantly shorter than the lifetime of the overall systems that they are built into, which poses great challenges of maintainability and sustainability. For long field-life systems, lifecycle mismatch between the system and its components has become one of the main costs. For instance, end-of-supply of spare parts for United States Navy systems has been estimated to cost up to 750 million dollars per year.
The main causes for ending supply of spare parts are technological developments and demand falls. Consequences can be mitigated by predicting, assessing, and actively managing end-of-supply risk. In this way, companies can decide to keep larger stock of parts that face ending supply but remain crucial for current business. Threat advisory systems for ending supply are very valuable, because it can be very expensive to find proper replacements at short notice. Therefore, evaluating end-of-supply risk of spare parts is the key factor in proactive management and strategic lifecycle planning for systems with long field-life.
Our research project presents a methodology to assess end-of-supply risk of spare parts using quantified upstream supply chain conditions. The methodology is developed from the perspective of purchasing firm, that is, firms that purchase parts, especially firms of long field-life systems. Such firms typically have only very limited access to the downstream supply chain information that is available to the parts manufacturers, such as parts sales data to perform lifecycle analysis. Indicators of end-of-supply risk are derived from information on the flow of spare parts from supplier to downstream companies in the supply chain. Four supply chain indicators are taken into account, that is, price and lead time that represent risks originating at the supply side, and cycle time and throughput that represent risks from the demand side. The methodology is demonstrated on data collected from a maintenance and repair organization in the aviation industry. Cross-validation results and out-of-sample risk assessment show good performance of our method. Further validation is provided by survey results obtained from the organization, which show strong agreement between the firm’s and our model’s identification of high-risk spare parts. All the results show that our method based on upstream supply information available to purchasing firms provides them with a helpful tool to reduce the risk of unforeseen ending supply of spare parts that are essential for their operation. Moreover, the joint incorporation of several risk indicators provides substantial gains over approaches based on a single risk indicator, stressing the importance of the joint analysis of big data.
Li, X., Dekker, R., Heij, C., & Hekimoğlu, M. (2016). Assessing End‐Of‐Supply Risk of Spare Parts Using the Proportional Hazard Model. Decision Sciences, 47(2), 373-394.