방법 비교

	Robust Event Tree Analysis	강건 고장 모드 및 영향 분석
분야	실험설계	실험설계
계열	Process / pipeline	Process / pipeline
기원 연도	1960s (ETA); robust extensions ~1990s–2000s	1980s–1990s
창시자	H.E. Lambert / Nuclear industry (ETA); robust extensions developed through aerospace and nuclear risk research	Extension of traditional FMEA (MIL-P-1629, 1949) integrated with Taguchi robust design philosophy (Genichi Taguchi, 1980s)
유형	Probabilistic risk assessment with uncertainty propagation	Risk analysis with variability quantification
원전	Bedford, T., & Cooke, R. M. (2001). Probabilistic Risk Analysis: Foundations and Methods. Cambridge University Press. ISBN: 9780521773201	Stamatis, D. H. (2003). Failure Mode and Effect Analysis: FMEA from Theory to Execution (2nd ed.). ASQ Quality Press. ISBN: 978-0873895989
별칭	Robust ETA, uncertainty-aware event tree analysis, ETA with uncertainty quantification, robust probabilistic event tree	Robust FMEA, Noise-Aware FMEA, Variability-Integrated FMEA, Robustness-Based FMEA
관련	6	4
요약	Robust Event Tree Analysis (Robust ETA) extends classical event tree analysis by explicitly accounting for uncertainty in the probability estimates assigned to each branch. Rather than treating branch probabilities as precise point values, the robust approach represents them as intervals, probability distributions, or imprecise probabilities, then propagates that uncertainty through the tree to produce outcome frequency ranges instead of single numbers. This gives decision-makers a clearer picture of the confidence in risk estimates under realistic conditions of incomplete or conflicting information.	Robust Failure Mode and Effects Analysis extends the classical FMEA framework by explicitly incorporating noise factors, parameter variability, and environmental variation into the risk assessment process. Rather than treating failure likelihood as a single deterministic estimate, it uses robust design principles — most notably from Taguchi's quality engineering — to evaluate how process variability and uncontrollable noise factors influence the probability and severity of each failure mode, yielding risk priority numbers that reflect real-world variability.