Winter Road Surface Condition Research Articles

Multimodal Transformer Model Using Time-Series Data to Classify Winter Road Surface Conditions

Multimodal Transformer Model Using Time-Series Data to Classify Winter Road Surface Conditions

Weather event characterization: a catalyst for improved spatial mapping and benefit quantification in winter road maintenance

Monitoring winter road surface conditions (RSC) is essential for optimizing winter road maintenance operations and ensuring public traffic safety and mobility. However, many existing RSC estimation methods focus primarily on local or site-specific conditions, thus lacking effectiveness in comprehensive spatial mapping under varying winter weather events. Additionally, the potential financial benefits these methods could provide to maintenance authorities remain largely unexplored. This study addresses these gaps by introducing a refined methodology that harnesses the K-Means algorithm to characterize various weather events – a significant stride towards improving the generalizability and accuracy of road surface temperature (RST) estimations. The study also incorporates regression kriging (RK), a renowned geostatistical technique, as an integral component of a system for constructing an extensive spatial map of RSC across large highway networks. This strategy maximizes the use of data from existing road weather information systems (RWIS) to bridge their large spatial gaps. This study further quantifies the potential savings derived from optimally placing RWIS stations and reducing traffic collisions. The efficacy of these methods is validated through a real-world case study on two major interstate highways in Iowa, US, with RST estimation discrepancies as low as 0.619 °C. The findings also indicate potential cost savings by conserving up to 10 RWIS stations, which can be further translated into monetary savings of RWIS capital costs, traffic collisions prevention, enhanced traffic mobility and savings in maintenance material usage.

Lightweight Regression Model with Prediction Interval Estimation for Computer Vision-based Winter Road Surface Condition Monitoring

Lightweight Regression Model with Prediction Interval Estimation for Computer Vision-based Winter Road Surface Condition Monitoring

Winter Road Surface Condition Recognition Using Semantic Segmentation and the Generative Adversarial Network: A Case Study of Iowa, U.S.A.

Adverse road surface conditions (RSCs) are a significant concern for traffic safety and mobility during the winter season. Dealing with this issue requires the prompt execution of maintenance operations to restore safe driving conditions. The faster maintenance personnel are made aware of the dangerous conditions, the faster they can mobilize to mitigate adverse RSCs. Therefore, the speed at which RSC information is provided is vital for reducing potential road-related incidents and maintaining mobility. Intending to improve the delivery of RSC information, this study proposes developing a deep learning-based method through the usage of road weather information system (RWIS) images to automate the RSC classification process. While the RWIS collects images that give a direct view of the road, the traditional manual way of utilizing these images for RSC monitoring is laborious and time-consuming. To overcome this challenge as well as other limitations associated with previously developed RSC recognition methods, the application of deep neural networks using imagery data is investigated here. More specifically, the application of semantic segmentation (SS) and generative adversarial network (GAN) techniques in automating RSC recognition was adopted in an attempt to predict the locations of drivable areas and indicate the snow hazard level solely based on RWIS images. Case study results from Iowa, U.S.A., show that both SS and GAN techniques perform their tasks with a high degree of accuracy.

Spatial Mapping of Winter Road Surface Conditions via Hybrid Geostatistical Techniques

Spatial Mapping of Winter Road Surface Conditions via Hybrid Geostatistical Techniques

An Automatic Architecture Designing Approach of Convolutional Neural Networks for Road Surface Conditions Image Recognition: Tradeoff between Accuracy and Efficiency

Convolutional neural network (CNN) is a promising image recognition technique for winter road surface condition (RSC), a measure that is crucial for winter maintenance operations. In the past, researchers have designed RSC CNN models that displayed acceptable results but did so focusing solely on obtaining high classification accuracy without any consideration for efficiency. Furthermore, when it comes to model development itself, architecture design requires expertise in CNN as well as rich knowledge in the investigated problem itself. To rectify these issues, this paper proposes an innovative approach to automatically design RSC CNN architecture without compromising classification accuracy. The proposed approach uses a weighted sum method, which provides the freedom of choosing relative importance level between accuracy and efficiency. Once the relative importance has been set, one of the most successful and widely adopted heuristics, namely, simulated annealing (SA), is employed to generate (sub)optimal solutions. Results show that both accuracy and efficiency of the automatically generated CNNs are better or at least comparable to the two selected state-of-the-art CNN models, ResNet50 and MobileNet, achieving as high as 93.44% classification accuracy. Ultimately, the outcome of this study fills the gap in existing CNN design methods that do not consider the tradeoff between accuracy and efficiency while providing insight into the effect varying architectures have on CNN model performance.

Winter road surface condition classification using convolutional neural network (CNN): visible light and thermal image fusion

During winter, road conditions play a crucial role in traffic flow efficiency and road safety. Icy, snowy, slushy, and wet road conditions reduce tire friction and affect vehicle stability which could lead to dangerous crashes. To keep traffic operations safe, cities spend a significant budget on winter maintenance operations such as snow plowing and spreading of salt and sand. This paper proposes a methodology for automated winter road surface condition classification using convolutional neural network (CNN) and the combination of thermal and visible light cameras. As part of this research, 4244 pairs of visible light and thermal images are captured from pavement surfaces and classified into snowy, icy, wet, and slushy surface conditions. Two single-stream CNN models (visible light and thermal streams) and one dual-stream CNN model are developed. The average F1-Score of dual-stream model is 0.866, 0.935, 0.985, and 0.888 on snowy, icy, wet, and slushy, respectively. The weighted average F1-Score is 0.94.

A Study on the Estimation of Winter Road Surface Conditions Using Measured Traffic Data

A Study on the Estimation of Winter Road Surface Conditions Using Measured Traffic Data

Real-time winter road surface condition monitoring using an improved residual CNN

This paper proposes a real-time winter road surface condition (RSC) monitoring solution that automatically generates descriptive RSC information in terms of snow and ice coverage by using images from fixed traffic and weather cameras. Several state-of-the-art pre-trained deep neural networks are customized and fine-tuned to address a specific domain, classifying the amount of snow coverage on a road surface. A thorough evaluation is conducted to identify and select the best model. This evaluation uses an extensive set of experiments to test the accuracy and generalization of each model and uses transfer-learning to fine-tune each of the pre-trained models on independent images from different traffic and weather cameras. The transferability of each model, relationship between model performance and data size, and the system settings of each model are then examined. Lastly, three online weight calibration methods are proposed to automatically update the model in new environments. The result shows that re-training the model using images from a mixed set of cameras has the most promising results.

An enhanced spatial statistical method for continuous monitoring of winter road surface conditions

To facilitate more efficient winter maintenance decision support, road weather information systems (RWIS) have been widely used by highway agencies. However, the cost of RWIS stations is high, and they have limited monitoring coverage. To address this challenge, this paper presents an innovative framework that applies regression kriging to integrate stationary and mobile RWIS data to improve the accuracy of road surface temperature (RST) estimation. Furthermore, an optimal RWIS network expansion strategy is introduced by incorporating a modified particle swarm optimization method with the objective of minimizing spatially averaged kriging estimation errors. A sensitivity analysis is also conducted to investigate the influence of station densities on model performance. The case study from Alberta, Canada, demonstrates the feasibility and applicability of the proposed method. The findings provide insights for continuous monitoring and visualization of both road weather and surface conditions and for optimizing RWIS network planning.

Use of Topography, Weather Zones, and Semivariogram Parameters to Optimize Road Weather Information System Station Density across Large Spatial Scales

A road weather information system (RWIS) is a combination of advanced technologies which collect, process, and disseminate road weather and condition information. This information is used by road maintenance authorities to make operative decisions that improve safety and mobility during inclement weather events. Many North American transportation agencies have invested millions of dollars to deploy RWIS stations to improve the monitoring coverage of winter road surface conditions. The design of these networks often varies by region, however, and it is not entirely clear how many stations are necessary to provide adequate monitoring coverage under different conditions; substantial gaps remain in knowledge about optimal design. To fill these gaps, an investigation was conducted to determine how optimized RWIS station densities relate to topographic and weather characteristics. A series of geostatistical semivariogram models were constructed and compared using topographic position index (TPI) and weather severity index (WSI) to measure relative topographic variation and weather severity, respectively. The geostatistical approach was then applied to map the optimum number of RWIS stations across several topographic and weather zones. The study area captured varying environmental characteristics, including regions with flat or varied terrain and warm or cold regions. This study suggests that RWIS data collected from a specific region can be used to estimate the number of stations required for regions with similar zonal characteristics. The outcome of this study can be used as a decision-making tool for RWIS network expansion, thus maximizing monitoring capability of RWIS networks using topographic and weather-related zonal classifications.

2C3-5 A Study of Winter Road Surface Condition Presentation by Steering Vibration

2C3-5 A Study of Winter Road Surface Condition Presentation by Steering Vibration

A Study on the Traffic Capacity Decrease Rate of Signal Intersection deu to Winter Road Surface Condition

A Study on the Traffic Capacity Decrease Rate of Signal Intersection deu to Winter Road Surface Condition

Does winter road maintenance help reduce air emissions and fuel consumption?

Winter road maintenance (WRM) has been shown to have significant benefits of improving road safety and reducing traffic delay caused by adverse weather conditions. It has also been suggested that WRM is also beneficial in terms of reducing vehicular air emissions and fuel consumptions because snow and ice on road surface often cause the drivers to reduce their vehicle speeds or to switch to high gears, thus decreasing fuel combustion efficiency. However, there has been very limited information about the underlying relationship, which is important for quantifying this particular benefit of a winter road maintenance program. This research is focused on establishing a quantitative relationship between winter road surface conditions and vehicular air emissions. Speed distribution models are developed for the selected Ontario highways using data from 22 road sites across the province of Ontario, Canada. The vehicular air emissions under different road surface conditions are calculated by coupling the speed models with the engine emission models integrated in the emission estimation model - MOVES. It was found that, on the average, a 10% improvement in road surface conditions could result in approximately 0.6–2% reduction in air emissions. Application of the proposed methodology is demonstrated through a case study to analyse the air emission and energy consumption effects under specific weather events.

Connected Vehicle Solution for Winter Road Surface Condition Monitoring

A connected vehicle–based winter road surface condition (RSC) monitoring solution that combines vehicle-based image data with data from a road weather information system is described. The proposed solution was intended as an improvement to a smartphone-based system evaluated in previous research. Three machine learning classification methods—namely, artificial neural networks, random trees, and random forests—were evaluated for potential application in the connected vehicle–based system for RSC monitoring. Field data collected during the 2014–2015 winter season were used for model calibration and validation. Results showed that all models improved the accuracy of the smartphone-based RSC classification substantially, with random forest having the highest classification performance. The models, however, were found to lack transferability; therefore, individual models will require local calibration before being used at any location.

Winter Road Surface Condition Monitoring

This paper describes the field evaluation of a smartphone-based winter road surface condition (RSC) monitoring system. The system was deployed on eight maintenance and patrol vehicles and tested during the winter of 2013–2014 on a section of Highway 6 in Ontario, Canada. The study focused on data obtained from four of the vehicles. This paper details the evaluation of the system from three perspectives: spot-wise monitoring, route-level monitoring, and system reliability. The study compared the system's RSC monitoring results with manual classifications, patrol records, and the Traveller's Road Information Portal, a public-facing website. The system's consistency with current RSC monitoring methods was assessed, and it was found that the smartphone-based system was capable of providing reliable results in comparison with the current method of patrol reporting for route-level monitoring of winter road conditions. Key performance factors were investigated, and the paper presents recommendations for system improvement.

Winter Road Surface Condition Forecasting

AbstractThis study has attempted to address a challenging problem in winter road maintenance, namely road surface condition (RSC) forecasting. A novel conceptual framework for short-term road surface condition forecasting is proposed. This framework is designed to consider all important conditional factors, including weather, traffic, and maintenance operations. Salt applications are modeled by considering a history instead of one single-time interval of salting operations. In this way, the variation of snow/ice melting speed caused by both residual salt amounts and salt-contaminant mixing state is effectively incorporated in the forecasting model, enabling accurate short-term forecasting for contaminant layers. This approach practically circumvents a major limitation of previous studies, making the postsalting RSC forecasting more reliable and accurate. Under this model framework, several advanced time series modeling methodologies are introduced into the analysis in order to capture the highly complex i...

Optimization of Vehicle Travel Position on Roundabouts in Snowy Cold Regions

Recently, modern roundabouts have been actively introduced in various countries. Single-lane roundabouts help to prevent serious accidents because vehicles tend to slow down as a result of the transverse acceleration that acts on them during movement on the circulatory roadway. However, as the boundary between the circulatory roadway and the truck apron becomes unclear during winter in snowy cold regions, it is necessary to verify the practicality of roundabouts in consideration of winter road surface and meteorological conditions. This study examined a method of guiding vehicles on circulatory roadways in winter. Driving experiments were conducted at a roundabout with different road surface conditions and central-island diameters to allow analysis of the positions of vehicles traveling under each set of conditions. The results indicated that vehicles were able to stay on course regardless of the central-island diameter in dry road surface conditions. With winter road surface conditions, however, the number of vehicles mounting the truck apron increased when the island diameter was small, whereas a large diameter helped them to stay on course and reduced travel position variance. An examination of driving behavior under different road surface conditions and with different central-island diameters revealed that surface conditions affected travel speed and transverse acceleration, whereas the island diameter had a smaller effect. However, it was found that the adoption of a large central-island diameter in a constant inscribed circle diameter was effective in reducing travel position deviation and helped vehicles to stay on the circulatory roadway even during the winter in snowy cold regions.

Relationship Between Winter Road-Surface Conditions and Vehicular Motion: Measurements by Probe Vehicles Equipped with Global Positioning System

A major concern of drivers in winter is the current road condition. Taxis, which move ceaselessly around a wide area, have great potential as sensors for detecting road-surface conditions in a given area. To establish a method with which to estimate road conditions based on the vehicular motion of taxis, field experiments were conducted by using probe vehicles fitted with vehicular-motion sensors and a Global Positioning System device, before implementation in taxis. Preliminary analyses were performed by using data measured on a test track, urban streets, and an expressway. The slip ratio, defined as the relative difference in speed between vehicle and tire wheel, was effective in indicating how slippery the road surfaces were. Taxi vehicular-motion data were collected for more than 1 month, although unlike with probe vehicles, the wheel speed was not measured. Some features of vehicular motion specific to slippery roads were identified, and the discriminability of road conditions, whether icy or dry, without the use of wheel-speed data, was examined.