Premium
Modeling infrastructure degradation from visual inspections using network‐scale state‐space models
Author(s) -
Hamida Zachary,
Goulet JamesA.
Publication year - 2020
Publication title -
structural control and health monitoring
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.587
H-Index - 62
eISSN - 1545-2263
pISSN - 1545-2255
DOI - 10.1002/stc.2582
Subject(s) - visual inspection , robustness (evolution) , data mining , scale (ratio) , computer science , visualization , structural health monitoring , reliability engineering , engineering , artificial intelligence , structural engineering , biochemistry , chemistry , physics , quantum mechanics , gene
Summary Visual inspections is a common approach for the network‐scale monitoring of bridges. One of the main challenges when interpreting visual inspections is the observations being subjective and thus the observation uncertainty varies among different inspectors. In addition, observations uncertainties can be dependent on the structural element condition. These two factors introduce difficulties in differentiating between measurement errors and legitimate changes in a structure's condition. This study proposes a state‐space model suited for the network‐scale analyses of transportation infrastructure. The formulation of the proposed framework enables quantifying the uncertainty associated with each inspector. In addition, the proposed model accounts for the uncertainty of visual inspections based on the structure condition as well as the uncertainty specific to each inspector. The predictive capacity and robustness of the proposed model are verified with synthetic inspection data, where the true deterioration state is known. Following the verification step, the proposed model is validated with real data taken from a visual inspections database.