Fault tree analysis is a well-known technique in reliability engineering and risk assessment that supports decision-making in complex systems. Traditional methods, however, require that parameters of fault trees (such as the failure rates of components) are precisely known.
In a recent paper, PrimaVera researchers Thom Badings, Nils Jansen, Matthias Volk, and Marielle Stoelinga propose a novel method that lifts this requirement. Their approach enables the analysis of fault trees, even if important parameters of the system are uncertain. The same applies to the more general model class of continuous-time Markov chains (CTMCs), of which fault trees are a specific instance.
A preprint of the full paper (to be presented this summer at the conference CAV), which is titled “Sampling-Based Verification of CTMCs with Uncertain Rates,” is available via this link.