Key points are not available for this paper at this time.
We present a Root Cause Analysis model for warranty failures that integrates the field failure data with the design improvement and internal quality-management data. Key features of the proposed model are: Ontology-Relationship-Diagram (ORD) to depict the Failure-Modes and Effect-Analysis (FMEA) data, conversion of ORD into a Bayesian Network (BN) in context to a Corrective Action Reports (CAR), usage of warranty claims data as the transactional data input to the BN to elicit probabilistic inference for warranty failure, application of text-processing technique for unstructured to structured data conversion. The benefits are reduction of Detection-To-Correction (DTC) cycle time and reduce liability exposure.
De et al. (Fri,) studied this question.