Automotive suppliers routinely analyse a large number of warranty claims for their products, for the purpose of design and process improvement, as well as warranty cost forecasting. In some instances, automotive warranty claims data are incomplete, which brings additional uncertainty to the process of analysis and forecasting. For example, it is not uncommon to have the vehicle build dates and warranty claim dates recorded, but the sale dates missing or unavailable to the supplier. This paper shows that using the build date instead of the sale date as the start of a mission life, provides drastic and sometimes unexpected differences in the outcomes of warranty claims forecasting. This paper presents a Monte Carlo simulation procedure to generate the missing time from build to sale data. It discusses different simulation algorithms, and analyses their accuracy, by comparing simulated data with actual data. In addition, the authors conducted a statistical analysis of build to sale data for various vehicle name plates and platforms, in order to enhance the forecasting procedure and improve its accuracy.
[1]
Roderick J. A. Little,et al.
Statistical Analysis with Missing Data
,
1988
.
[2]
Mark P. Kaminskiy,et al.
G-renewal process as a model for statistical warranty claim prediction
,
2000,
Annual Reliability and Maintainability Symposium. 2000 Proceedings. International Symposium on Product Quality and Integrity (Cat. No.00CH37055).
[3]
D. Murthy,et al.
New product warranty: A literature review
,
2002
.
[4]
B. Singh,et al.
Customer-Rush Near Warranty Expiration Limit, and Nonparametric Hazard Rate Estimation From Known Mileage Accumulation Rates
,
2006,
IEEE Transactions on Reliability.
[5]
Andre Kleyner,et al.
Minimizing life cycle cost by managing product reliability via validation plan and warranty return cost
,
2008
.
[6]
Ming-Wei Lu,et al.
Automotive reliability prediction based on early field failure warranty data
,
1998
.
[7]
Asit P. Basu,et al.
Statistical Methods for the Reliability of Repairable Systems
,
2000
.