Artificial Intelligence tools to predict the level of defectiveness of existing bridges

The “Italian guidelines for the maintenance of bridges” propose a qualitative method for the classification of structural, seismic, hydraulic and geotechnical risk of infrastructures. Focusing on the structural risk, one of the key parameters that significantly drives the evaluation of the vulnerability class is the level of defectiveness. The level of defectiveness can be determined only after the execution of the visual inspections, which are necessary to point out the type of damages that affect the structure, their intensity, size and position in each structural component of the bridge. Given the high number of structures to be checked and the time that is necessary to execute the visual inspections of all these bridges, an instrument to have a starting idea of the conservation status of the structure could be helpful to establish an order of priority for the bridges to be investigated. With such purpose in mind, this paper presents an ongoing activity based on the use of Artificial Intelligence to develop a smart tool that recognizes the different elements that compose the bridge, the defects, their intensity, size and position. This tool could be applied to an automatic image collection process, e.g. using a drone that, with minimal user interaction, captures images of the structure and provides quick feedback to the operators.