Bayesian Surrogates for Risk-Aware Pre-Assessment of Aging Bridge Portfolios

TL;DR

The paper tackles the problem of quickly and reliably pre-assessing aging bridge portfolios to decide where detailed, costly analyses are necessary. It introduces a parametric NLFEA data-generation pipeline and trains three Bayesian Neural Networks to predict code-compliance factors $\\boldsymbol{\\eta}(\\mathbf{x})$ with calibrated epistemic uncertainty, enabling fast, uncertainty-aware triage. Key contributions include the data-generation workflow, calibrated BNN surrogates for each compliance factor, and a portfolio-level triage policy that guides refinement decisions; the approach supports reduced-input deployment via SHAP-informed feature ranking. In a real-world railway underpass case, the method demonstrates accurate predictions and calibrated uncertainties, enabling avoidance of unnecessary interventions and substantial cost and emissions savings. Overall, the work offers a scalable, uncertainty-aware screening framework for aging infrastructure that can markedly reduce unnecessary assessments while directing resources to high-risk structures.

Abstract

Aging infrastructure portfolios pose a critical resource allocation challenge: deciding which structures require intervention and which can safely remain in service. Structural assessments must balance the trade-off between cheaper, conservative analysis methods and accurate but costly simulations that do not scale portfolio-wide. We propose Bayesian neural network (BNN) surrogates for rapid structural pre-assessment of worldwide common bridge types, such as reinforced concrete frame bridges. Trained on a large-scale database of non-linear finite element analyses generated via a parametric pipeline and developed based on the Swiss Federal Railway's bridge portfolio, the models accurately and efficiently estimate high-fidelity structural analysis results by predicting code compliance factors with calibrated epistemic uncertainty. Our BNN surrogate enables fast, uncertainty-aware triage: flagging likely critical structures and providing guidance where refined analysis is pertinent. We demonstrate the framework's effectiveness in a real-world case study of a railway underpass, showing its potential to significantly reduce costs and emissions by avoiding unnecessary analyses and physical interventions across entire infrastructure portfolios.

Figures (2)

• Figure 1: Graphical user interface of the developed ML-based structural pre-assessment tool for reinforced concrete frame bridges. Users input structural parameters (left) and receive rapid code compliance predictions with uncertainty estimates (right).
• Figure 2: Predictive performance in the safety-critical region $\eta \in [0.5,1.5]$. CB reported after post-hoc scaling.