Fault Trees Explain "Why It Went Wrong" Using Causality

Fault trees are widely used risk models that help identify potential failures in complex systems, primarily through minimal cut set analysis which answers "what can go wrong?". This research extends the utility of fault trees by integrating them with Halpern & Pearl's theory of actual causality. This integration allows fault trees to address the crucial question of "why has it gone wrong?", making them a powerful tool for failure diagnostics. The study offers a comprehensive classification of various notions of actual causality within the context of fault trees. This classification is based on both the graphical and logical structures inherent in fault trees. Furthermore, the research demonstrates a clear relationship between minimal cut sets, which represent the smallest combinations of events that can lead to a system failure, and the identification of actual causes for that failure. This advancement provides a more profound understanding of system failures, moving beyond mere identification of potential issues to pinpointing the specific causal factors.