Asset management of gravel roads: Mobile mapping and data science approach
Loading...
URL
Journal Title
Journal ISSN
Volume Title
School of Engineering |
Master's thesis
Unless otherwise stated, all rights belong to the author. You may download, display and print this publication for Your own personal use. Commercial use is prohibited.
Authors
Date
2024-12-30
Department
Major/Subject
Spatial Planning and Transportation Engineering
Mcode
Degree programme
Master's Programme in Spatial Planning and Transportation Engineering
Language
en
Pages
41
Series
Abstract
Gravel roads play a vital role in maintaining connectivity and supporting economic activities, particularly in rural areas. However, the assessment and maintenance of these roads face challenges due to their vulnerability to environmental conditions, such as erosion, rutting, and seasonal changes. Traditional manual inspections, commonly used for gravel road management, are labour-intensive, costly, and prone to errors, particularly when handling large-scale networks. Additionally, existing research in road geometry analysis often focuses on paved roads, leaving gravel roads underexplored. This creates a critical gap in developing automated, efficient, and reliable methods for detecting geometric anomalies specific to gravel roads, which are essential for ensuring road safety and serviceability. This thesis proposes an innovative approach to address this gap by integrating advanced mobile mapping techniques with data science algorithms. Using a laser scanning system, high-resolution road profile data were collected across 39,301 meters of gravel roads in Finland. The algorithm employs B-spline interpolation to reduce noise and smooth the road profiles, with gradient estimation and additional numerical tricks to detect key anomalies such as intersections, road widening, and structural features like bridges. Independent analysis of the left and right sides of the road enhances the algorithm’s ability to handle asymmetric features, while convolution-based smoothing ensures resilience to noise and irregularities in the dataset. The methodology was designed to address the limitations of traditional approaches by offering an automated and adaptive solution. The results demonstrate the robustness of the algorithm, with accurate detection of intersections, road widening, and bridges. This research addresses the limitations of traditional manual inspections by offering an automated, scalable solution. The findings contribute to proactive gravel road management strategies, aligning with Finland’s National Transport System Plan (2021–2032) to enhance road maintenance efficiency and sustainability. Future research will focus on improving label accuracy, addressing complex scenarios such as overlapping anomalies, and expanding the dataset to include varied geographic conditions, making this approach adaptable to broader global applications.Description
Supervisor
Mladenovic, MilosThesis advisor
Wang, HaishanMakowska, Michalina
Keywords
gravel roads, anomaly detection, mobile mapping, road geometry analysis, asset management, slope analysis