Achieving quality assurance in accepting automated pavement condition data: threshold development and error pattern analysis

Automated pavement condition survey systems are widely used for network-wide pavement distress detection, yet concerns about data quality remain. To address this, it is crucial to develop effective data quality assurance (QA) procedures. This study proposes two indexes for QA procedures: the accurac...

Full description

Saved in:
Bibliographic Details
Published in:The international journal of pavement engineering 2024-12, Vol.25 (1)
Main Authors: Tao, Jueqiang, Luo, Xiaohua, Wang, Feng, Faieq, Ajmain, Gong, Haitao, Qiu, Xin
Format: Article
Language:English
Subjects:
acceptance threshold
Automation
Data analysis
difference of two-sigma
Error analysis
Error detection
image data
Image quality
Network reliability
Pattern analysis
Pavement condition
Pavements
Quality assurance
Quality control
Thresholds
Citations: Items that this one cites
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Automated pavement condition survey systems are widely used for network-wide pavement distress detection, yet concerns about data quality remain. To address this, it is crucial to develop effective data quality assurance (QA) procedures. This study proposes two indexes for QA procedures: the accuracy index based on the differences between automated measurements and manual audit measurements, and the precision index based on the differences between automated measurements of two consecutive years. Using automated distress data (2017-2021) and manual audit data from Texas, the Difference of Two-Sigma (D2S) method was applied to determine the thresholds of various distress types for the accuracy and precision indexes. Then, a validation of the proposed thresholds was conducted by visually checking the 2022 image data of a selected pavement network. Through this process, the reliability of the thresholds was evaluated, and the error patterns present in the automated data were analysed. The study results demonstrated that the developed thresholds were able to effectively identify data quality issues. The primary source of errors in the automated condition survey system was false-positive detections of pavement distresses, frequently caused by features in the pavement surface background, primarily due to the textures of the pavements.
ISSN:1029-8436
1477-268X
DOI:10.1080/10298436.2024.2433614