|
Activity Number:
|
30
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, July 29, 2007 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Quality and Productivity
|
| Abstract - #309836 |
|
Title:
|
Monitoring of Hazard Rates for Dynamically Changing Observations
|
|
Author(s):
|
Emmanuel Yashchin*+
|
|
Companies:
|
IBM T.J. Watson Research Center
|
|
Address:
|
, Yorktown Heights, NY, 10598,
|
|
Keywords:
|
Warranty ; Reliability ; Control Charts ; Cusum ; Change Point ; Wearout
|
|
Abstract:
|
We consider life testing situations in which a relatively large portion of data that serves as a basis for a monitoring scheme changes with time. Situations of this type are common in the areas of accelerated life testing or warranty data analysis. One of the key problems in such situations is detection of changes in the hazard rate for the tested population. For example, in the warranty data setting one may want to detect, as quickly as possible, onset of excessive wearout conditions for some set of manufacturing vintages. Since the whole data set is affected at each point in time that new information is obtained, one needs a special approach to the problem of monitoring lifetime characteristics. In this paper we discuss one such approach and give an example related to the problem of detection of trends in warranty data for electronic components.
|
