|
Independent Evaluation Concludes Meridium Software Provides State-of-the-Art Methods for Reliability Analysis
[ 5/13/2009 ]
Roanoke, Virginia, USA – Meridium Inc., the leader in asset performance management (APM) software and consulting solutions, today announced that an independent evaluation of Meridium algorithms incorporated into the reliability analysis routines within Meridium software were found to provide accurate implementation of state-of-the-art methods for reliability estimation. The independent verification effort, which took slightly over one year to complete, was conducted by Dr. Joel A. Nachlas, Associate Professor, Department of Industrial and Systems Engineering, Virginia Polytechnic Institute and State University, USA. Some key reported findings were: - algorithms incorporated in Meridium reliability analysis software are among the most up-to-date and widely accepted methods for reliability analysis;
- algorithms being used are well understood and appear to be correctly implemented in the Meridium code; and
- data analysis results provided by the Meridium code conformed to or matched, as appropriate, the results obtained using independently constructed code and should therefore be considered to be statistically accurate.
Software performance investigation included routines that treat the most commonly used distribution models for equipment reliability and specifically examined software accuracy in relation to various levels of data censoring. For each distribution, the software performance in computing both point estimates and confidence intervals on the parameters as well as confidence levels on distribution parameters and quantiles were evaluated. Supporting computations of Fisher information statistics, goodness of fit measures and maximum likelihood functions were included in the evaluation. After identification of the precise statistical algorithms and numerical analysis techniques included in Meridium software, separate independent computer code was constructed. Several test data sets were constructed for each of the distribution models and the data sets were analyzed using Meridium code and the separate independent code. The analysis was performed for complete and censored data sets where censoring was general and the censored fraction was varied over the range of 40% to 75%. “Meridium products are the result of the unparalleled asset performance management software development expertise and experience of our staff. Our success is ultimately tied to the success of our customers. Going the extra mile to seek independent and unbiased verification that our products are state-of-the-art – the best the market has to offer – is just one more level of confidence we can provide our customers,” says Bonz Hart, Meridium founder and CEO. ________________ About Dr. Joel A. Nachlas Dr. Nachlas’ teaching and research within the area of Industrial Operations and Systems Engineering spans 38 years. He has authored and co-authored over 28 articles, published three books and been the principal research investigator in numerous research activities including work for TRW, Inc., IBM Corporation, the National Science Foundation and the U.S. Department of Energy. His primary research focus is the development of new analytical strategies for application of probability theory and statistical methods to problems in reliability, maintenance planning and quality control. In addition to his current position as Associate Professor in the Department of Industrial and Systems Engineering at Virginia Polytechnic Institute and State University, USA, Dr. Nachalas’ other places of academic employment include the University of Pittsburgh, USA; the University of Angers, France; and the University of Nice, France. Dr. Nachlas received a B.E.S. in Industrial Engineering and Operations Research from Johns Hopkins University and a M.S. and Ph.D. in Industrial Engineering and Operations Research from the University of Pittsburgh.
|
