Reliability Modeling and Experimental Design for the Extended Range Cannon Artillery Increment 2

Reliability is an essential element of system suitability for defense weapon systems. It often takes a prominent role in both the design and analysis of developmental and operational tests. However, in the current era of reducing budgets and increasing reliability requirements, it is increasingly important to incorporate all available information to effectively understand a system and design efficient tests. Bayesian approaches to reliability assessment provide formal statistical models to incorporate heterogeneous information and assess uncertainty. This effort will focus on reliability modeling for the Extended Range Cannon Artillery Increment 2 (ERCA-II) long-range precision fires platform. Previous work at USMA has elicited information from subject-matter experts and developed an initial component-by-component beta-binomial reliability model with a supporting R Shiny implementation application. This proposal extends the initial work to address three goals. The first goal is to extend the current modeling to incorporate relevant historical information. Guo and Wilson (2013) provide details of how subject-matter expertise, pass-fail data, lifetime data, and degradation data can be integrated into a full Bayesian assessment of system reliability. The goal of incorporating heterogeneous data from both the system and its components is to provide a simultaneous assessment of system and component reliability and uncertainty. In addition, these extensions will consider models that incorporate dependent data. A key use of the reliability modeling is for designing subsequent experimentation. A traditional reliability demonstration test is essentially a classical hypothesis test, which uses only the data from the current test to assess whether the reliability-related quantity of interest meets or exceeds a requirement. Many modern systems are highly reliable and extremely complex. For these systems, reliability demonstration tests often require an impractical amount of testing. In response to this dilemma, one can develop reliability assurance tests that use additional supplementary data and information to reduce the required amount of testing. The additional data and information may include appropriate reliability models, earlier test results on the same or similar devices, expert judgment regarding performance, knowledge of the environmental conditions under which the devices are used, benchmark design information on similar devices, prior knowledge of possible failure modes, etc. The developing of this kind of reliability assurance testing strategy requires principled methods for combining information. Work to date from USMA has developed an initial R Shiny application to incorporate subject-matter expertise into beta distributions. The third goal will focus on integrating the additional modeling from and making a more robust transferable application. This research and application will use state-of-the-art approaches to analyze ERCA-II reliability and understand the uncertainties in the estimates of system and component reliability. Integrating all available data will allow the development of efficient test plans and the evaluation of the risks and tradeoffs of those plans.