Probabilistic digital twin for continuous bridge performance modeling

The multiplicity of possible different and possibly interacting processes contributing or governing deterioration processes in bridges requires systems representations way beyond the deterministic approach where the focus is on few processes and corresponding models. A Probabilistic Digital Twin (PDT), supported by Structural Health Monitoring (SHM) and big data from observations and monitoring, can provide a novel contribution to support integrity management with knowledge improved over time. This paper explores the preliminary analyses for the future full development of the PDT as part of the Bridgitise Project. An approach to build the PDT of a bridge will be developed using FEM modelling of the deterioration effects. The simulated structural responses under deterioration progressing in time will be used to develop a PDT model that will then be utilized as a representation of the best available knowledge and applied to generate simulations of very significant numbers of structural performance scenarios over time.