This paper aims to deal with multiple data sets from different levels of complex on-demand systems. The paper will propose a method for incorporating overlapping higher level and lower level data in a Bayesian construct in order to update component reliability information. The technique can then be used to allow coordinated evidence sets from various system levels to reveal as much information as possible, and hence allow sensor placement optimization.
The expert elicitation technique discussed in this paper conveys a method of risk-informed design performed in support of NASA Lunar Surface Systems design that is guided by system design documents and based heavily on face-to-face designer interaction and elicitation. This approach has proven to be very efficient, as designers are closely engaged early in design cycles and forced to focus on reliability strategies that were heavily influenced and implemented by the designer’s own expertise.
This paper presents a new method to address the likelihood estimates bias as a result of small sample size and the new distribution attributes and flexibility.
This paper presents a new methodology combining the quantitative and qualitative Bayesian Belief Networks together to do the risk assessment and reliability analysis.