Reliability Analysis and Prediction with Warranty Data: Issues, Strategies, and Methods

Through simple, practical approaches, Reliability Analysis and Prediction with Warranty Data: Issues, Strategies, and Methods helps Six Sigma black belts and engineers successfully interpret warranty data to make accurate predictions. It discusses how to use this data to define and analyze field problems, provides guidelines for discovering the root causes for warranty cost reduction, and explores issues associated with warranty data and the approaches to overcome them.

The first part of the book presents an introduction to reliability analysis and prediction using warranty data and highlights the issues involved. The second section offers strategies and methods for obtaining component-level nonparametric hazard rate estimates that provide important clues toward probable root causes and that help reduce warranty costs. Focusing on the prediction of warranty performance, the final part deals with methodologies that assess the impact of changes in warranty limits and forecast warranty performance.

This user-friendly book shows how warranty data can support various levels of decision making to achieve reliable outcomes. Easily understood even for those with minimal statistical background, it includes objectives and summaries in each chapter to enable quick review of the topics.

Table of Contents

NEED FOR ANALYSIS AND PREDICTION WITH WARRANTY DATA AND ISSUES INVOLVED

Reliability Studies with Warranty Data: Need and Issues

Continuous Improvement and Field Data

Three Levels of Decision Making With Warranty Data

The Role of Hazard Function in Reliability and Robustness

Improvements

Warranty Data: Not Always Perfect for Statistical Analysis

Existing Research Work and Certain Limitations

The Scope and Objective of the Book

Warranty Concepts

Organization of the Book

Bibliographic Notes

Characterization of Warranty Data

Life Cycle of a Vehicle

An Overview of the Warranty Claim Process

Automobile Warranty Data: Key Characteristics

Two Inherent Characteristics of Warranty Data: Uncleanliness and Incompleteness

Summary

STRATEGIES AND METHODS FOR RELIABILITY ANALYSIS WITH WARRANTY DATA

Strategies for Reliability Analysis from Warranty Data

From Customer Concerns to Root Causes

Strategies for Hazard Function Estimation

Summary

Hard Failures with Known Mileage Accumulation Rates

Risk Set Adjustment in Hazard Function

Modeling of Mileage Accumulation Rate in the Vehicle Population

A Four-Step Methodology

An Application Example

Summary

Hard Failures with Unknown Mileage Accumulation Rates

Hazard Function with Modification in Numerator

Modeling Mileage on Failed Vehicles Using Truncated Normal Distribution

A Five-Step Methodology

An Application Example

Summary

Bibliographic Notes

Soft Failures with Known Mileage Accumulation Rates

Introduction

Risk Set Adjustment: MIS as Life Variable

Risk Set Adjustment: Mileage as Life Variable

Incorporating Censoring Information in the Hazard Function Estimation

A Six-Step Methodology

An Application Example

Summary

Bibliographic Notes

Soft Failures with Unknown Mileage Accumulation Rates

Hazard Function

Estimation from Doubly Truncated Data Sets

Incorporating Censoring Information in the Hazard Function Estimation

A Seven-Step Methodology

An Application Example

Summary

WARRANTY PREDICTION

Assessing the Impact of New Time/Mileage Warranty Limits

Changes in the Warranty Coverage

The Estimation of Number of Warranty Claims

Cost of Warranty Claims

Summary

Bibliographic Notes

Forecasting of Automobile Warranty Performance

Warranty Growth or Maturing Data Phenomena

Current Warranty Forecasting Methods

Warranty Performance Forecasting Using RBF Neural Networks

Forecasting Warranty Performance Using MLP Neural Networks

Results and Discussions

Summary

Bibliographic Notes

References

Index

Objectives and Overview sections appear at the beginning of each chapter.

Write a review