State-Space Models for Network-Scale Analysis of Bridge Inspection Data

Abstract

Visual inspection is one of the main techniques for monitoring the deterioration in networks of bridges. Visual inspection data commonly display inconsistencies due to the subjective nature of the evaluation and because different individuals perform the inspections over time. One of the main challenges when interpreting visual inspections is differentiating between measurement errors and legitimate changes in a structures condition. This study proposes a state-space model for modeling the deterioration behaviour while accommodating the inspector-induced uncertainty. The proposed framework allows modeling inspections uncertainty according to the current structures state as well as the variability associated with each inspector. The predictive capacity of the proposed framework is verified with synthetic inspection data, where the true deterioration state is known.