Domenic Di Francesco

The project is developing a generic Bayesian decision analysis framework to quantify the expected value of information from inspections on a structure. The influence of selected system effects on the analysis is being investigated. These include:

Dependencies in degradation modelling

Fatigue is the primary damage mechanism that is considered. Bayesian data analysis is being used to fit predictive models that coherently and consistently account for variability within and dependency between model parameters. In addition, the effect of dependency between uncertain inputs is also being considered.

Improved characterisation of imperfect information

The project is developing improved methods of characterising inspection activities in terms of the precision, bias and relevance of the information that they provide, as well as the risks and expected costs of completion. This will allow a more comprehensive evaluation regarding the extent to which they facilitate improved risk management strategy.

Spatial dependencies

Where dependencies can be quantified between structural condition at different locations on a structure, the project will quantify the additional value of information that results from any inference (updating) that can be completed at locations other than those directly inspected.