An Integrated Approach for Automated Acquisition of Bridge Data and Deficiency Evaluation

Bridges are essential components of civil infrastructures that enable joining highways and cities, as well as facilitating movement between different geographical locations. Deterioration of bridges can pose a threat to public safety and hinder economic activities. The development of innovative and reliable asset management tools to support the optimization of maintenance and timely intervention plans is of paramount concern to transportation agencies. However, such development is usually impeded by a lack of data and/or difficulties in collecting and organizing data from different sources. This paper introduces an integrated scheme for automated acquisition of bridge-related data. The proposed methodology enables the creation of a unified structured and comprehensive data repository that can be used for further processing and modeling purposes. Data from multiple sources, including web pages and bridge inspection reports, were extracted, collected, prepared, and compiled. As such, web scraping and rule-based information extraction were employed to achieve this objective. Finally, text mining was applied to the inspection comments extracted from bridge inspection reports to get insights into such inspection information and evaluate the severity of bridge deck deficiencies.