Neural Network Approach to Condition Assessment of Highway Culverts: Case Study in Ohio

    Authors

    O. Tatari; S. M. Sargand; T. Masada;B. Tarawneh

    Comments

    Authors: contact us about adding a copy of your work at STARS@ucf.edu

    Abbreviated Journal Title

    J. Infrastruct. Syst.

    Keywords

    Culverts; Risk assessment; Neural networks; Inspections; Highways; RISK-ASSESSMENT; PIPE; INSPECTION; MANAGEMENT; Engineering, Civil

    Abstract

    Millions of culverts exist in the United States, and they are aging rapidly. Inspection of all the culverts consumes a lot of time and resources. Instead of inspecting each culvert every 5years, this study presents a more intelligent approach to predict the condition of each culvert. An artificial neural network (ANN) model is built to assess the condition of the culverts based on culvert inventory data. The overall condition-rating predictions are compared with the condition rating based on manual inspection. The results of this study have shown that ANN was able to predict culvert adjusted overall rating with high precision, as the course of action score prediction rate was 100%. Sensitivity analysis of the ANN model is provided to assess the effect of variables. The goal of this study is to show that more intelligent culvert-management systems could be devised by taking advantage of artificial intelligence. (C) 2013 American Society of Civil Engineers.

    Journal Title

    Journal of Infrastructure Systems

    Volume

    19

    Issue/Number

    4

    Publication Date

    1-1-2013

    Document Type

    Article

    Language

    English

    First Page

    409

    Last Page

    414

    WOS Identifier

    WOS:000329913100006

    ISSN

    1076-0342

    Share

    COinS