A spatial, stochastic, chloride induced corrosion model of reinforced concrete structures is briefly analyzed in this work. Modeling the spatial variability of model parameters gives one the ability, not only to quantify the probability of degradation, but the extent of damage as well. The model aspires to predict structures' extent of damage for their whole service life period. An efficient way of predicting reliable results, throughout the years, is by utilizing inspection information to update the model. This stochastic inverse problem, concerning the spatial extent of damage, presents many challenges however, in realistic applications, mainly due to computational intractability. To overcome this obstacle, a series of steps are taken in this work consisting of sensitivity analysis, quadrature techniques and the unscented transform. The specific updating procedure allows the solution of the problem in reasonable computational time, even for large structures and sophisticated numerical models.
