Tracking types and extent of pavement deterioration is critical for maintaining road networks in a serviceable condition. The prevailing methods for obtaining pavement condition data include manual and semiautomated surveys, which are time-consuming and involve significant human intervention. Extensive research has been performed in automating the process for more efficient, objective, and repeatable distress evaluations. This paper highlights the preliminary results from an effort sponsored by the Florida Department of Transportation to develop and implement automated software for identification and quantification of pavement surface cracking distresses. A technical framework was developed for systematic evaluation of available automated technologies in contrast to manual methods. Pertinent performance measures were identified to evaluate the accuracy, precision, repeatability, reproducibility, and efficiency of various methods. This framework was implemented to determine the gaps in effectiveness of automated applications, to design corresponding solutions, and to gauge reliability expectations accordingly. The evaluation follows two main steps: ( a) comparison of the cumulative quantities of various distress types found in manual surveys versus automated surveys and ( b) verification of the automatically detected distresses against reference crack maps generated through a semiautomated process of manually rating the collected images. Although the overall comparison of distress quantities indicates strengths and weaknesses of the evaluated algorithm, the distress by distress verification of software performance is used to identify design solutions for addressing the indicated weaknesses. The guidelines in this systematic framework can be modified with context-sensitive considerations to be applicable to other highway agencies transitioning to automated applications.
Evaluation Framework for Automated Pavement Distress Identification and Quantification Applications
Transportation Research Record
Transportation Research Record: Journal of the Transportation Research Board ; 2639 , 1 ; 46-54
2017-01-01
Article (Journal)
Electronic Resource
English
Automated Pavement Distress Analyzer
British Library Conference Proceedings | 1994
|Automated Real-Time Pavement Distress Analysis
Transportation Research Record | 1999
|Automated Real-Time Pavement Distress Analysis
Online Contents | 1999
|Transportation Research Record | 2008
|