An enhanced framework for dynamic segmentation of pavement sections Conference Paper uri icon

abstract

  • © 2017 by Canadian Society for Civil Engineering. All rights reserved. Over the past decades, highway agencies have used automated and semi-automated data collection methods such as laser scanning and ultrasonic waves, resulting in the collection of an enormous amount of high-density pavement condition data. The agencies are now able to quantify the level of extent and severity of different distresses even for extremely short length of pavement segments. A scientific method to aggregate those small pavement sections into a reasonable size of segments plays an important role in accurately representing the overall pavement network performance and making practical maintenance and rehabilitation decisions. This paper reviews the current methodologies for segmenting pavement condition data and summarizes their limitations. Then, the study proposes a new segmentation framework for pavement sections that finds homogenous segments by considering multiple pavement distresses for performance evaluation and treatment selection purposes using the affinity propagation clustering technique and heuristic rules. The affinity propagation clustering technique finds the similarity between pavement sections based on the distress data. However, the clustering technique does not consider the spatial nature of pavement features. As such, heuristic rules are formulated to overcome this limitation and identify homogenous pavement segments. A case study is conducted to illustrate the capabilities and applications of the proposed segmentation framework. The proposed segmentation framework will improve the a) representation of pavement condition data, b) formulation of pavement maintenance and rehabilitation strategies, and c) pavement performance evaluation.

published proceedings

  • 6th CSCE-CRC International Construction Specialty Conference 2017 - Held as Part of the Canadian Society for Civil Engineering Annual Conference and General Meeting 2017

author list (cited authors)

  • Abdelaty, A., & Jeong, H. D.

complete list of authors

  • Abdelaty, A||Jeong, HD

publication date

  • January 2017