Title: Using Machine Learning for Condition Assessment of Concrete Infrastructure

Author(s): Leandro F.M. Sanchez and Marcelo Terra

Publication: Concrete International

Volume: 41

Issue: 11

Appears on pages(s): 35-39

Keywords: damage, image, structure, data

DOI:

Date: 11/1/2019

Abstract:
The damage rating index (DRI) is among the most promising microscopy methods for condition assessment of concrete infrastructures. However, it requires experts to perform time-consuming petrographic microscopy of samples. The article discusses a study aimed at automating DRI assessments using machine learning (ML). It is expected that ML will reduce the subjectively yet increase the speed, reproducibility, accessibility, and accuracy of DRI determination.