TL;DR
This paper introduces a non-contact method using 3D laser scanning and a friction tester to evaluate sidewalk skid resistance. By analyzing surface morphology (texture, joints) and correlating it with friction measurements under dry and wet conditions, the study provides a framework for improving pedestrian and cyclist safety and enabling proactive sidewalk management.
What Problem Does This Paper Address?
Sidewalks are critical infrastructure, but injuries due to slips and falls are common, especially with increasing use of e-bikes and an aging population. Existing methods for assessing sidewalk safety, particularly skid resistance, often rely on manual, contact-based measurements that can be time-consuming, labor-intensive, and subject to human variability. Furthermore, these methods may not fully capture the complex interplay of factors like surface texture, joint design, and drainage that contribute to skid resistance. The paper aims to bridge this gap by developing a non-contact, automated approach to comprehensively evaluate sidewalk safety.
Methods
The study employs a two-pronged approach:
- Friction Measurement: A T2GO continuous friction coefficient tester was used to measure skid resistance under both dry and wet conditions. This device is described as having high reliability and adhering to ASTM and EN standards.
- 3D Surface Morphology Analysis: A Gocator 2350 laser scanner was used to capture high-resolution 3D surface data of the sidewalks. This allowed for the extraction of various morphological metrics, including:
- Mean Texture Depth (MTD): Represents the depth of the surface texture.
- Fractal Dimension (FD): Quantifies the horizontal complexity of the surface texture.
- Macro/Micro-texture Density (HLTx and WLTx): Distinguishes between larger-scale (macro) and smaller-scale (micro) texture features.
- Joint Dimensions: Including seam width, depth, and density.
These metrics were then analyzed using hierarchical regression analysis. This method allowed the researchers to assess the incremental contribution of different factors (texture vs. joint parameters) to skid resistance. Variance Inflation Factor (VIF) was used to check for multicollinearity among independent variables. Model performance was evaluated using R², ΔR², and Mean Absolute Error (MAE).
Datasets
The study utilized data collected from ten different sidewalk sites in Shanghai, China. These sites represented a variety of common sidewalk paving types, including permeable bricks, stone, and tiles. The researchers collected both 3D surface morphology data and friction coefficient measurements for these sites under both dry and wet conditions. No pre-existing datasets were used for training or evaluation; the data was collected specifically for this study.
Key Insights and Ideas
- Dry vs. Wet Skid Resistance Drivers: The study highlights a significant difference in the primary drivers of skid resistance between dry and wet conditions. While texture complexity (FD) and overall texture depth (MTD) are key for dry conditions, micro-texture density (WLTx) and joint depth become more critical for wet conditions.
- The Double-Edged Sword of Texture: High fractal dimension (FD > 2.5) is beneficial for dry skid resistance, indicating complex textures that enhance grip. However, very high macro-texture density (HLTx > 80) can be detrimental in wet conditions, potentially hindering drainage. Similarly, dense micro-textures (high WLTx) can impede water drainage, reducing wet skid resistance.
- Joints Matter for Drainage: Seam density and depth play a crucial role in wet skid resistance, primarily by facilitating drainage. However, excessive joint density can negatively impact friction. A balanced design is key.
- Non-Contact Scanning is Powerful: The research validates that non-contact laser scanning is an efficient and effective method for capturing detailed surface texture and joint data, which is crucial for understanding skid resistance mechanisms.
Key Results
- Hierarchical Regression Models:
- Dry Skid Resistance: The model
Du_Dry = 0.61·FD – 0.53·MTD + 0.21·Depth + εindicates that FD positively influences dry friction, while MTD negatively influences it. Seam depth also shows a slight positive contribution.
- Dry Skid Resistance: The model
- Wet Skid Resistance: The model
Du_Wet = -0.72·WLTX +0.31·Depth + εshows that micro-texture strength (WLTx) is a strong negative predictor of wet friction, while depth positively contributes, likely due to drainage. - Performance Metrics: The models achieved good performance, with R² values of 0.72 for dry and 0.85 for wet conditions in the initial texture-only stage, and improved to 0.75 and 0.89 respectively after adding seam parameters. MAE values were 0.08 for dry and 0.05 for wet, indicating manageable prediction accuracy.
- Material Performance: Certain materials, like stones #4 and #5, showed excellent dry and wet skid resistance. Conversely, some tiles (#6, #7) exhibited low wet friction coefficients, falling below national standards.
- Joint Design Recommendations: Seam density should ideally be controlled below 5%. For wet environments, prioritizing materials with low WLTx (<3500) and sufficient joint depth (≥0.25 mm) is recommended.
Limitations and Future Work
- Limitations:
- The study acknowledges that the availability of data was limited, which influenced the choice of statistical models (OLS instead of more complex hierarchical models).
- The testing efficiency of existing high-precision laser scanners might limit their scalability for large-scale urban data collection.
- The study focused on specific sidewalk types in Shanghai; generalizability to other regions or materials might require further validation.
- Future Work:
- Refining predictive models through multi-sensor fusion is suggested.
- Developing standardized design guidelines for sidewalks.
- Expanding the dataset and refining methodologies to enhance the robustness of findings.
- Optimizing sensor deployment and data extraction processes for improved efficiency and scalability.
- Exploring the feasibility of shared paved spaces for bikes/wheelchairs through pavement contact modeling using finite element analysis.
Summary updated