Put distributions on node parameters (epistemic/parameter uncertainty) and propagate to system-reliability bounds / credible intervals — distinct from the aleatory Monte-Carlo already hardened.
The array-vectorised system_probability already supports Monte-Carlo over parameter draws, so the substrate exists.
Estimate
Effort: M (2–4 days) · Value: Med–High
The vectorised system_probability is already the substrate, so the work is an API for parameter-level uncertainty plus the draw/summarise loop. The aleatory-vs-epistemic distinction is increasingly expected of a serious reliability tool, which lifts this above a pure nice-to-have.
Put distributions on node parameters (epistemic/parameter uncertainty) and propagate to system-reliability bounds / credible intervals — distinct from the aleatory Monte-Carlo already hardened.
The array-vectorised
system_probabilityalready supports Monte-Carlo over parameter draws, so the substrate exists.Estimate
Effort: M (2–4 days) · Value: Med–High
The vectorised
system_probabilityis already the substrate, so the work is an API for parameter-level uncertainty plus the draw/summarise loop. The aleatory-vs-epistemic distinction is increasingly expected of a serious reliability tool, which lifts this above a pure nice-to-have.