@article{10902/38557,
year = {2025},
month = {11},
url = {https://hdl.handle.net/10902/38557},
abstract = {A method for conducting Bayesian elicitation and learning in risk assessment is presented. It assumes that the risk process can be described as a fault tree. This is viewed as a belief network, for which prior distributions on primary event probabilities are elicited by means of a pairwise comparison approach. A novel and fully Bayesian updating procedure, following different observation campaigns of the events in the fault tree for the posterior probabilities assessment, is described. In particular, the goal is to handle contexts where there are limited data information (one of the challenges for elicitation), thus keeping simple the elicitation process and adequately quantifying the uncertainties in the analysis. Often, an important consideration in these contexts is the trade-off between how many of the events in the fault tree can be observed against the information that the extra data yield. How this can be addressed within this method is demonstrated. The application is illustrated through three real examples, including the motivating example of risk assessment of spacecraft explosion during controlled reentry.},
organization = {This research is partially supported by the Insight Centre for Data Analytics, funded by Science Foundation Ireland through grant 12/RC/2289-P2, and by the Spanish Ministry of Science, Innovation and Universities through grant PID2022-137818OB-I00. It is also partially supported by the NPI program of the European Space Agency (ESA) and by the Spanish Ministry of Science, Innovation and Universities, under the program “Salvador de Madariaga,” grant PRX18/00128. We are immensely grateful to Dr. Guillermo Ortega of ESA, who provided insight and expertise that greatly assisted the research. We would like to thank those at the ESA who gave their time and expertise to the elicitation process for the ATV application.},
publisher = {Wiley-Blackwell},
publisher = {Risk Analysis, 2025, 45(11), 4014-4038},
title = {Quantitative system risk assessment from incomplete data with belief networks and pairwise comparison elicitation},
author = {De Persis, Cristina and Bosque Orero, José Luis and Huertas, Irene and Sillero-Denamiel, M. Remedios and Wilson, Simon P.},
}