Root Cause Analysis of Product Service Failure Using Computer Experimentation Technique

Abstract

Improvement of product quality and reliability following field failures of products is a well-researched area. The feedback from the service in the form warranty/field performance is important in guiding design and manufacturing changes. Traditionally, efforts in this area have been limited to analysis of warranty data for evaluating warranty policies, estimating field reliability, predicting failures etc. Also some research has been done on performing corrective actions in manufacturing via process adjustments by linking warranty failures with manufacturing Key Control Characteristics (KCCs). However there is also need to develop the systematic approach of doing Root Cause Analysis (RCA) and Corrective Actions in design. This research proposes a novel methodology of performing RCA and CA in design by linking warranty failures with the product design parameters such as Geometric Dimensioning and Tolerance (GD&T). An analytical approach based on computer experimentation technique performs effective Root Cause Analysis (RCA) of product failures by linking warranty failure modes with the design parameters and identifies the analytical relationship between them. The method focuses on identifying root cause(s) to address in tolerance product design faults. Warranty failures modes are represented by Key Failure Characteristics (KFCs) and the geometrical design parameters are represented by Key Product Characteristics (KPCs). The proposed methodology establishes the analytical relationship between the KFCs and KPCs using variation mechanistic models to identify the dimensional variations which caused the failure. A case study on automotive ignition switch demonstrates the methodology.