Validation and Implementation of AHTD Network-Level Automated Distress Rating and Pavement Condition Rating

摘要

Pavement conditions are commonly described by the amount,severity,and frequency of pavement distresses.Transportation engineers prefer to measure,as accurately as they are able,cracking,rutting,and pavement profile for pavement condition ratings.These measurements are most often used exclusively for pavement condition assessment.The use of cracking,rutting,and ride quality (a function of pavement profile) to produce a value representing the pavement's overall condition is followed in great detail by engineers and is well documented in the literatures [1,2].Automated systems designed to assess network-level pavement distress for rutting and ride quality have been developed and widely utilized.These automated systems have the accuracy and precision in most cases without concern.However,the use of cracking data to complete the pavement condition index has lagged behind because of the time necessary to manually process cracking data and/or the high variability of automated systems.While automated pavement distress surveys are more efficient,the quality of the automated crack survey data has been highly questionable since its conception.This paper summarizes the implementation,validation,and results of a network-level automated crack survey in Arkansas.The study showed that the performance of a two-dimensional (2D) automated distress survey largely relies on the pixel intensity of the images,which could be affected by the pavement aggregate type,pavement age,the environment (moisture and temperature),and lighting condition under which the images were collected.To investigate how these factors affected the automated cracking survey data,a sensitivity study was conducted at the project-level for a variety of environmental and lighting conditions.It was found that noise (false-positive) is an inevitable obstacle in the automated crack survey considering the broad varieties of image characteristics.Careful design of cracking protocols supplemented with proper validation process can overcome some of the limitations of the automated survey.The study demonstrated the efficiency of the automated survey to assist a state transportation agency's pavement management activities.At the same time,it also exposed the limitations of the 2D automated cracking survey.A discussion of utilizing 3D technology for automated cracking survey is included which highlights its great potential in automated pavement condition assessment.