PROJECT TITLE :

Adaptive Warranty Prediction for Highly Reliable Products

ABSTRACT:

Field come rate prediction is very important for manufacturers to assess the product reliability and develop effective warranty management. To urge timely predictions, lab reliability tests have been widely employed in assessing field performance before the product is introduced to the market. This work issues warranty prediction for highly reliable products. However, because of the high reliability related to trendy electronic devices, the failure knowledge in lab tests are typically insufficient for each individual product, resulting in less correct prediction for the field return rate. To overcome this issue, a hierarchical reliability model is instructed to efficiently integrate the knowledge from multiple devices of an analogous sort in the historical database. Underneath a Bayesian framework, the warranty prediction for a brand new product can be inferred and updated as the information collection progresses. The proposed methodology is applied to a case study in the information and Communication technology business for illustration. Bayesian prediction is demonstrated to be terribly effective compared to different alternatives via a cross-validation study. In specific, the prediction error rate based mostly on our updating prediction theme is significantly improved as more field data are collected, and achieves a prediction error rate less than 20p.c after launching the merchandise for 3 months.