- Leandro Iannacone [0000-0002-1946-3940], University College London, l.iannacone@ucl.ac.uk
- Paolo Gardoni, University of Illinois Urbana-Champaign, gardoni@illinois.edu
Abstract
Deterioration can cause severe reductions in the performance of complex infrastructure systems. The decrease in performance can affect both regular service and the ability to respond to disastrous events. When modeling a deteriorating infrastructure, several sources of uncertainty might cause hard-to-predict behaviors. Stochastic models are required to obtain the evolution of the system with the associated uncertainties, and the recent literature has seen a rise in stochastic process theory applied to the field of deteriorating infrastructure. The most recent advancements range from physics-based models that use Stochastic Differential Equations for the evolution of the state variables of the system, to empirical methods that leverage the potential of artificial intelligence and machine learning. In both cases, the stochastic methods come with unknown model forms and parameters that must be calibrated based on field data from Non-Destructive Testing (NDT) or Structural Health Monitoring (SHM). The field data must be integrated within the assumed mathematical models to achieve a realistic representation of the system response. In the context of infrastructure analysis, a better understanding of the conditions of critical components (and of the deterioration processes that alter those conditions over time) can help to devise optimal maintenance strategies and to accurately assess and reduce the disruptions that natural and man-made hazards might cause.
This mini-symposium aims to bring together expert researchers and academics concerned with the various aspects of infrastructure subject to deterioration. In particular, the session welcomes studies that focus (i) on novel stochastic models for deterioration analysis and quantification of the associated uncertainties, (ii) on the collection of relevant data via NDT/SHM, and (iii) on the calibration of deterioration models to predict the future performance of the system based on the data. The contributions could range from applications of Machine-Learning and Artificial Intelligence for the calibration of stochastic models, to theoretical contributions in the fields of stochastic processes for deterioration modeling.
Keywords: Deterioration, Reliability Analysis, Infrastructure Analysis, Non Destructive Testing, Structural Health Monitoring, System Identification
