module has either or both attribute and time-to-failure components.The raw data for assessments are the component failure history data and the system configuration.The historical data are "successes and failures" for binomial-Bernoulli components and "failures and testing time (normalized to 'mission equivalent units')" for time-to-failure components.The configuration data consist of a list or lists of minimal paths ("minimal path sets" or "tie sets"), or else a list of minimal cuts ("minimal cut sets"), for the system as a list of modules, and for each module as a list of components. If the MTBF assessment option is selected, the system "mission time" is also needed.The underlying mathematical model, identical with that incorporated into the first version of the SPARCS program described in [5], is an amalgamation of Boolean logic, probability, and Bayesian and Monte Carlo techniques.The system reliability, a numerical-valued function of the component reliabilities, is derived by the method of inclusion-exclusion (IE), also known as Poincar•'s theorem, from the minimal paths or the minimal cuts. The failure-history data are "sufficient statistics," for the parameters of Bayesian conjugate prior distributions (c.p.d.'s) on the component reliabilities, "beta" for attributes and "negative-log gamma" for time to failure. Since each of these divisions deals with a separate aspect of the work performed under the contract, they are presented herein as two separate papers, with independent lists of references. The introduction presents the historical background for this project, including prior work directly related to software development, input data documentation, and an example. Division 1 is a broad-range tutorial on system reliability analysis, covering a very broad range of related subject matter. Division 2 is specialized to a discussion of the beta and gamma random deviate generators incorporated into the software developed under this project and documented in the companion "SPARCS-2 Users Manual," AFFDL-TR-78-18, Volume II.This work was performed under work unit 2304N104, System ReliabilityConfidence Assessment, with Dr. H. Leon Harter as project engineer.iii The raw data for assessments are the component failure history data and the system configuration. The historical data are "successes and failures" for binomial-Bernoulli components and "failures and testing time (normalized to 'mission equivalent units')" for time-to-failure components.The configuration data consist of a list or lists of minimal paths ("minimal path sets" or "tie sets"), or else a list of minimal cuts ("minimal cut sets"), for the system as a list of modules, and for each module as a list of components. If the MTBF assessment option is selected, the system "mission time" is also needed. this made it possible to generate the equation and calculate the probability for a system of substantially larger sizes than can be processed with SCOPE.3. "Minimal path" or else "minimal cut" calculations: by taking advantage of the dual relationship betwe...
