Pavement Missing Condition Data Imputation through Collective Learning-Based Graph Neural Networks
arXiv:2603.06625v1 Announce Type: new Abstract: Pavement condition data is important in providing information regarding the current state of the road network and in determining the needs of maintenance and rehabilitation treatments. However, the condition data is often incomplete due to various reasons such as sensor errors and non-periodic inspection schedules. Missing data, especially data missing systematically, presents loss of information, reduces statistical power, and introduces biased assessment. Existing methods in dealing with missing data usually discard entire data points with missing values or impute through data correlation. In this paper, we used a collective learning-based Graph Convolutional Networks, which integrates both features of adjacent sections and dependencies between observed section conditions to learn missing condition values. Unlike other variants of graph neural networks, the proposed approach is able to capture dependent relationship between the conditi
arXiv:2603.06625v1 Announce Type: new Abstract: Pavement condition data is important in providing information regarding the current state of the road network and in determining the needs of maintenance and rehabilitation treatments. However, the condition data is often incomplete due to various reasons such as sensor errors and non-periodic inspection schedules. Missing data, especially data missing systematically, presents loss of information, reduces statistical power, and introduces biased assessment. Existing methods in dealing with missing data usually discard entire data points with missing values or impute through data correlation. In this paper, we used a collective learning-based Graph Convolutional Networks, which integrates both features of adjacent sections and dependencies between observed section conditions to learn missing condition values. Unlike other variants of graph neural networks, the proposed approach is able to capture dependent relationship between the conditions of adjacent pavement sections. In the case study, pavement condition data collected from Texas Department of Transportation Austin District were used. Experiments show that the proposed model was able to produce promising results in imputing the missing data.
Executive Summary
This paper proposes a novel approach to impute missing pavement condition data using a collective learning-based Graph Convolutional Networks (GCNs). The proposed model leverages both the features of adjacent sections and dependencies between observed section conditions to learn missing condition values. The approach is tested on pavement condition data collected from the Texas Department of Transportation Austin District, demonstrating promising results in imputing missing data. This research has significant implications for pavement maintenance and rehabilitation treatments, as accurate and complete condition data is essential for making informed decisions. The proposed method addresses the limitations of existing approaches, which often discard entire data points with missing values or impute through data correlation. By capturing dependent relationships between adjacent pavement sections, the proposed model provides a more accurate and realistic representation of pavement conditions.
Key Points
- ▸ The proposed method uses a collective learning-based GCNs to impute missing pavement condition data.
- ▸ The approach leverages both features of adjacent sections and dependencies between observed section conditions.
- ▸ The model is tested on pavement condition data collected from the Texas Department of Transportation Austin District.
Merits
Strength in Handling Dependent Relationships
The proposed method effectively captures dependent relationships between adjacent pavement sections, providing a more accurate and realistic representation of pavement conditions.
Improved Data Imputation Results
The approach demonstrates promising results in imputing missing data, outperforming existing methods that discard entire data points or impute through data correlation.
Demerits
Limited Scalability
The proposed method may not be scalable to large datasets, as the computational complexity of GCNs can increase exponentially with the size of the input graph.
Data Quality Requirements
The approach assumes that the input data is of high quality and relevance, and may not perform well with noisy or irrelevant data.
Expert Commentary
The proposed method represents a significant advancement in the field of missing data imputation, particularly in the context of transportation infrastructure management. By leveraging the strengths of GCNs, the approach effectively captures dependent relationships between adjacent pavement sections, providing a more accurate and realistic representation of pavement conditions. However, the method's limitations, such as its limited scalability and data quality requirements, must be carefully considered. Future research should aim to address these limitations and explore the application of the proposed method in other transportation-related tasks.
Recommendations
- ✓ Further research should be conducted to investigate the scalability of the proposed method and explore strategies for improving its performance on large datasets.
- ✓ The approach should be tested on a variety of transportation infrastructure management tasks to demonstrate its applicability and effectiveness.