The research objective of this award is to establish a new paradigm of multidisciplinary system reliability prediction that enables the use of evolving, insufficient data sets (from expert knowledge, customer survey, system inspection & testing, and field data) over the entire product life-cycle. Three technical tasks will drive this new paradigm: (1) A Bayesian Information Toolkit (BIT) will be built to precisely model evolving data sets (soft and hard data; input and output data); (2) A Bayesian reliability method for component failure modes will be developed for problems with evolving, insufficient data sets, and; (3) A Bayesian system reliability method will be established, which requires a rigorous understanding of statistical coupling among multiple failure modes. The proposed research plan will be carried out through collaborative efforts with the General Motors (GM) R&D Center, the U.S. Army TARDEC, and Whirlpool Corporation.
If successful, this research will fill the substantial gap between theory and practice in system realization. The proposed multidisciplinary system reliability prediction tools will advance engineering analysts' approaches for system realization improving techniques from the current state of heuristic and qualitative methods to the level of systematic-collaborative methods. In the future, in addition to strategic validation, this capability will substantially increase the role of simulation technology in product design because simulations can replicate uncertain nature in complex system (e.g., vehicle, airplane) operation. Upon completion of the proposed research, the PI will disseminate research materials (e.g., unclassified uncertainty data, case study examples) to academic researchers for their research enhancement and commercialize research assets (e.g., tools for multidisciplinary system reliability prediction) with a software vendor. This research will become mission-critical to the US economy because it will substantially reduce the current effort ($53 billion/year in the automotive industry alone) required to fix defects from maintenance and post-quality care.
