Software for Use of Expert Opinion in Reliability


Aboura, Khalid; Singpurwalla, Nozer D.; Soyer, Refik   (1989, ASQC)   George Washington University, Washington, DC

Annual Quality Congress, Toronto, Ontario, Canada    Vol. 43    No. 0
QICID: 3616    May 1989    pp. 527-532
List $10.00
Member $5.00

This article is not available online. Contact us to receive a scan of the archive, in PDF format.

Article Abstract

The use of expert opinion or informed judgment is becoming prevalent in practical applications of reliability and risk analysis. The issue is particularly aggravated by the fact that modern day components and systems are designed for high reliability and so the availability of even a small amount of failure data is difficult in many situations. In current practice the treatment of expert opinion is undertaken in an ad hoc manner, usually with the experts sitting around the table and reaching a consensus through discussion. Although their results are expressed quantitatively, the approach does not take into account the experts' biases and differing degrees of uncertainty regarding the numbers given. Also, the method of combining differences of opinions is not formalized.

In this paper we discuss an interactive PC-Based software for reliability assessment of items whose life lengths are described by a Weibull distribution. We demonstrate a software named "IPRA" that was developed at the Institute for Reliability and Risk Analysis, the George Washington University. The novel feature of the program pertains to the incorporation of informed judgment or expert opinion into the analysis, and the provision for incorporating the analyst's opinions on the expertise of the experts. The code is user friendly and leads a reliability analyst through the process of eliciting expert opinion on the deterioration/improvement rate and the median life of the items. The program also allows use of failure and/or survival data on identical items and provides an interactive and time-sequential merging of the modulated expert opinion with the data. It provides tabular and graphical displays of prior and posterior reliability functions and reliability interval distributions for different mission times. "IPRA" also provides feedback to the expert, and in the light of new data, the analyst can call upon the program to update the results.



Browse QIC Articles Chronologically:     Previous Article     Next Article

New Search

Featured advertisers

ASQ is a global community of people passionate about quality, who use the tools, their ideas and expertise to make our world work better. ASQ: The Global Voice of Quality.