Controlled experiments comparing fault-tree-based safety analysis techniques

Abstract

The capability to model dynamic aspects of safety-critical systems, such as sequence or stochastic dependence of events, is one important requirement for safety analysis techniques. State Event Fault Tree Analysis, Dynamic Fault Tree Analyis, and Fault Tree Analysis combined with Markov Chains Analysis have been developed to fulfill these requirements, but they are still not widely accepted and used in practice. In order to investigate the reasons behind this low usage, we conducted two controlled experiments. The goal of the experiments was to analyze and compare applicability and efficiency in State Event Fault Tree analysis versus Dynamic Fault Tree Analyis and Fault Tree Analysis combined with Markov Chains Analysis. The results of both experiments show that, notwithstanding the power of State Event Fault Tree Analysis, Dynamic Fault Tree Analyis is rated by participants as more applicable and is more efficient compared to State Event Fault Tree Analysis, which, in turn, is rated as more applicable but is less efficient than Fault Tree Analysis combined with Markov Chains Analysis. Two of the reasons investigated are the complexity of the notations used and the lack of tool support. Based on these results, we suggest strategies for enhancing State Event Fault Tree Analysis to overcome its weaknesses and increase its applicability and efficiency in modeling dynamic aspects of safety-critical systems.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 6 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.