Project Intelligent Road is a common effort of four main partners: Rovaniemi University of Applied Sciences (RAMK), Luleå Technical University (LTU), Finnish Meteorological Institute (FMI), Kaakkois-Suomen ELY-keskus (KaS ELY) – and representatives of business sector of the region. The overall objective of the project is to support the development of business community in Northern Scandinavian region by testing and improvement of existing innovative products concerned with road safety provision in Nordic weather conditions. The specific objective of the project is to create a demo of sustainable and marketable Intelligent Road System providing location-based short-term road weather information to the road user passing by the area. Target Groups, which are directly and positively affected by the project are businesses and above mentioned project partners. All these actors have exclusive know-how in the area and are each equally important for the implementation of the project and its success. Final beneficiaries of this project are business community of the Northern Scandinavian Region; Ministry of Transport and Communications of Finland, represented by KaS ELY and Finnish Transport Safety Agency; Swedish Ministry of Enterprise, Energy and Communications; and of course local municipalities, as Luleå Municipality, and road users.

Snow is a heterogenous, hot material which is constituted from ice particles. The bonding behavior of ice particles is an important parameter determining the macroscopic behavior of snow. Discrete Element Method (DEM) is usually used as a tool to model dry snow. The most important input data required into the DEM is bonding behavior of ice particles since ice particles can adhere to form bonds when they brought into contact. This study had two aims: first, an analytical formulation was derived to predict the bond diameter of ice-ice contacts as a function of time, compressive load, and strain rate. Using the previously published data for strain rate of ice, a solution method was developed. The results of bond diameter development with time were compared to experimental data and a good agreement was found. Second, a DEM for dry snow was developed and programmed in MATLAB and the developed bond model was employed in the simulation to study the deposition behavior of snow in a container under gravity acceleration. A specific beam element with implemented damage model was developed in implemented in the simulation using the bond data obtained from the analytical approach. The simulated parameters were macroscopic angle of repose, packing density, and surface conditions as a function of temperature and filling rate. The results showed that discrete element simulations were able to verify the existing published experimental data. Specifically, the simulation results showed that angle of repose of snow decreased rapidly with decreasing the temperature, the surface became very irregular due to the particles rotation and re-arrangement for lower falling speeds of particles, and density increased with depth of deposition. These findings were all matched with experimental observations.

Despite the significance of snow in various cryospheric, polar, and construction contexts, more comprehensive studies are required on its mechanical properties. In recent years, the utilization of μ CT has yielded valuable insights into snow analysis. Our objective is to establish a methodology for mapping density-dependent material properties for dry manufactured snow within the density range of 400–600 kg/m 3 utilizing μ CT imaging and step-wise, quasi-static, mechanical loading. We also aim to investigate the variations in the structural parameters of snow during loading. The three-dimensional (3D) structure of snow is captured using μ CT with 801 projections at the beginning of the experiments and at the end of each loading step. The sample is compressed at a temperature of − 18 o C using a constant rate of deformation (0.2 mm/min) in multiple steps. The relative density of the snow is determined at each load step using binary image segmentation. It varies from 0.44 in the beginning to nearly 0.65 at the end of the loading, which corresponds to a density range of 400–600 kg/m 3 . The estimated modulus and viscosity terms, obtained from the Burger’s model, show an increasing trend with density. The values of the Maxwell and Kelvin–Voigt moduli were found to range from 60 to 320 MPa and from 6 to 40 MPa, respectively. Meanwhile, the viscosity values for the Maxwell and Kelvin–Voigt models varied from 0.4 to 3.5 GPa-s, and 0.3–3.2 GPa-s, respectively, within the considered density range. In addition, Digital Volume Correlation (DVC) was used to calculate the full-field strain distribution in the specimen at each load step. The image analysis results show that, the particle size and specific surface area (SSA) do not change significantly within the studied range of loading and densities, while the sphericity of the particles is increased. The grain diameter ranges from approximately 100 μ m to nearly 400 μ m, with a mode of nearly 200 μ m. The methodology presented in this study opens up a path for an extensive statistical analysis of the material properties by experimenting more snow samples.

We experimentally demonstrate the presence of a capillary bridge in the contact between an ice particle and a smooth aluminum surface at a relative humidity of approximately 50% and temperatures below the melting point. We conduct the experiments in a freezer with a controlled temperature and consider the mechanical instability of the bridge upon separation of the ice particle from the aluminum surface at a constant speed. We observe that a liquid bridge forms, and this formation becomes more pronounced as the temperature approaches the melting point. We also show that the separation distance is proportional to the cube root of the volume of the bridge. We hypothesize that the volume of the liquid bridge can be used to provide a rough estimate of the thickness of the liquid layer on the ice particle since in the absence of other driving mechanisms, some of the liquid on the surface must have been pulled to the bridge area. We show that the estimated value lies within the range previously reported in the literature. With these assumptions, the estimated thickness of the liquid layer decreases from nearly 56 nm at T = −1.7°C to 0.2 nm at T = −12.7°C. The dependence can be approximated with a power law, proportional to (TM − T)−β, where β < 2.6 and TM is the melting temperature. We further observe that for a rough surface, the capillary bridge formation in the considered experimental conditions vanishes.

We present experiments along with an approximate, semi-analytic, close-form solution to predict ice sintering force as a function of temperature, contact load, contact duration, and particle size during the primary stage of sintering. The ice sintering force increases nearly linear with increasing contact load but nonlinear with both contact duration and particle size in the form of a power law. The exponent of the power law for size dependence is around the value predicted by general sintering theory. The temperature dependence of the sintering force is also nonlinear and follows the Arrhenius equation. At temperatures closer to the melting point, a liquid bridge is observed upon the separation of the contacted ice particles. We also find that the ratio of ultimate tensile strength of ice to the axial stress concentration factor in tension is an important factor in determining the sintering force, and a value of nearly 1.1 MPa can best catch the sintering force of ice in different conditions. We find that the activation energy is around 41.4KJ/mol41.4KJ/mol, which is close to the previously reported data. Also, our results suggest that smaller particles are “stickier” than larger particles. Moreover, during the formation of the ice particles, cavitation and surface cracking is observed which can be one of the sources for the variations observed in the measured ice sintering force.

The mechanical properties of natural snow play a crucial role in understanding glaciers, avalanches, polar regions, and snow-related constructions. Research has concentrated on how the mechanical properties of snow vary, primarily with its density; the integration of cutting-edge techniques like micro-tomography with traditional loading methods can enhance our comprehension of these properties in natural snow. This study employs CT imaging and uniaxial compression tests, along with the Digital Volume Correlation (DVC) to investigate the density-dependent material properties of natural snow. The data from two snow samples, one initially non-compressed (test 1) and the other initially compressed (test 2), were fed into Burger’s viscoelastic model to estimate the material properties. CT imaging with 801 projections captures the three-dimensional structure of the snow initially and after each loading step at -18C, using a constant deformation rate (0.2 mm/min). The relative density of the snow, ranging from 0.175 to 0.39 (equivalent to 160–360 kg/m), is determined at each load step through binary image segmentation. Modulus and viscosity terms, estimated from Burger’s model, exhibit a density-dependent increase. Maxwell and Kelvin–Voigt moduli range from 0.5 to 14 MPa and 0.1 to 0.8 MPa, respectively. Viscosity values for the Maxwell and Kelvin–Voigt models vary from 0.2 to 2.9 GPa-s and 0.2 to 2.3 GPa-s within the considered density range, showing an exponent between 3 and 4 when represented as power functions. Initial grain characteristics for tests 1 and 2, obtained through image segmentation, reveal an average Specific Surface Area (SSA) of around 55 1/mm and 40 1/mm, respectively. The full-field strain distribution in the specimen at each load step is calculated using the DVC, highlighting strong strain localization indicative of non-homogeneous behavior in natural snow. These findings not only contribute to our understanding of natural snow mechanics but also hold implications for applications in fields such as glacier dynamics and avalanche prediction.

This thesis presents a technology and method for classifying and characterizing different road conditions such as dry asphalt and asphalt covered with water, ice and snow. The method uses light sources of different wavelengths to illuminate the road surface and a detector to measure the reflected light from the road surfaces, dependent on how the surface have absorbed, scattered and polarized the light it is possible to classify the road condition. However, knowing what substrate that is on the asphalt is not enough to make good road grip estimations. Hence, by applying a radiative transfer model and estimate parameters such as the porosity, roughness and depth of the substrate it is possible to get more information that could improve a road grip estimate. Such investigations were carried out both in a laboratory environment and on actual roads.The technology here presented shows potential for classifying and characterizing different road conditions. Statistics shows that many traffic accidents with fatal outcome can be related to slippery road conditions. The most hazardous road conditions are the ones that are hard for the driver to detect and that appears sudden on the road. A sensor that can classify road conditions, could improve road safety considerably. Road condition information could be used in several applications, such as in the own vehicle both to inform the driver and safety applications such as the electronic stability program (ESP), anti-lock brake system (ABS) or the traction control system (TCS) on the prevailing road conditions. Such information could also be sent to other drivers and to the infrastructure, making it possible to plan a travel route depending on the prevailing road conditions. Or as information to road maintenance so they know where work is needed. Since there are a number of prototypes and ideas how to estimate the road grip, it is important that different methods are evaluated and tested in the same way. Hence, this thesis includes test methods and metrics to verify the various systems and investigate the weaknesses and strengths of the technologies. This is also carried out for the technology presented in this thesis.

Statistics shows that most traffic accidents with fatal outcome can be related to slippery road conditions. The most hazardous road conditions are the ones that are hard for the driver to detect and that appears sudden on the road. A sensor that classify the road condition in front of the vehicle, warning both the driver and the systems in the vehicle that are incorporated to help the driver, like the electronic stability program (ESP), anti-lock brake system (ABS) or the traction control system (TCS), could help to reduce these accidents. There are several prototypes for classification of road conditions available but they are not yet fully functional. In this thesis a method that makes it possible to classify the four distinct road conditions dry asphalt and asphalt covered with water, ice and snow, respectively, with a low probability of wrong classification using three wavelengths is presented. A prototype sensor built on the a technique using two laser diodes and a photo detector is tested in a real environment and compared with laboratory measurements which shows a promising result characterizing dry asphalt and asphalt covered with ice and snow. Both theory and experiments are presented. The most difficult road conditions to classify from each other are water and clear ice for which a method using polarized light is investigated. The investigation shows that using polarized light for illumination and a polarizer as an analyzer for classification of water and ice on asphalt is a more reliable method than using unpolarized light. All three investigations show promising results in developing an actual sensor to reduce fatal accidents in traffic.

Driver awareness of current winter road conditions (RCs) is known to affect the frequency of accidents due to sudden changes in these conditions. For example, partially icy roads that appear during autumn in northern areas typically result in collisions and ditch runs unless the drivers are generally aware of the situation. Availing motorists who drive under winter RCs of enhanced information is therefore highly desirable to increase their awareness of hazardous driving conditions. Such conditions need to be predicted ahead of time and presented to drivers before they attempt slippery road sections. Moreover, the identification of slippery RCs should quickly trigger targeted road maintenance to reduce the risk of accidents. This study presents a scalable and reusable collaborative intelligent transport system, herein referred to as an RC information system (RCIS). RCIS provides accurate RC predictions and forecasts based on RC measurements, road weather observations, and short-term weather forecasts. The prediction methods in the context of the distributed RCIS have been tested using a prototype implementation. These tests confirmed that these inputs could be combined into useful and accurate information about winter RCs that can be adapted for different types of users.

Today, there are a lot of vehicles and tyre testing carried out on lake ice surfaces. Thus, it is important to have knowledge about parameters that affect roadgrip. The thesis within this paper is that the liquid like layer which appears due to increasing temperature can be reduced by manipulating the ice roughness. This in turn should decrease the temperature dependence of the roadgrip in temperatures around 0°C. In order to investigate this, measurements of temperature, surface roughness and hardness and roadgrip were performed on three outdoor ice surfaces using an IR thermometer, an optical sensor with three IR-diodes, a steel ball drop indentation test and an RT3 curve, respectively. Additional ice roughness measurements were also made on two tempered ice surfaces in an ice hall. Results show a clear connection between ice temperature and roadgrip, unfortunately the created ice roughness was too small to influence the change in roadgrip

Road friction measurement is an important issue for active safety systems on vehicles; hence knowledge of this key parameter can significantly improve the interventions on vehicle dynamics. This study compares two different on-board sensors for the classification of road conditions with polarised infrared light. Several tests are performed on a dedicated track, with focus on detection of dry or wet surfaces, and the presence of ice or snow. The work shows the capability of both sensors to provide a correct classification. In particular, results indicate how the monitored area, the presence of active illumination and the mounting position influence measurements and response times. It is concluded that both systems classify different road conditions in all cases. Performance of the RoadEye system varied from 80 to 90 % whereas the camera based IcOR achieved 70-80 % accuracy level. Since these are being prototype sensors more development is needed before implemented into advanced safety applications.

Low adhesion between railway wheel and rail, usually induced by contaminants such as water, oil, and leaves, affects railway performance and safety. This study uses an optical sensor to identify various surface layers that cause low adhesion. In both a laboratory set-up and field tests under various conditions, the surface layers were identified by the optical sensor; in addition, the friction coefficients for the same surface layers were measured. The results indicate that the sensor can distinguish between different surface layers. This information, together with friction coefficient data, can be used in a prediction system for use by railway operators.

In order to classify the road condition, dry asphalt and asphalt covered with water, ice and snow a technique using a sensor called Road eye is presented. The Road eye sensor uses three wavelengths and one photo detector to determine the intensities that are reflected from the road surface and is then able to estimate the road condition. By linking the Road eye sensor to a GPS and a Mulle, a miniature wireless Embedded Internet System, the road conditions can be associated with the correct road position, making it possible to use the information in many different applications.

Roadgrip is an important parameter for vehicle testing and road maintenance. Therefore, an evaluation of the velocity and curvature effects on roadgrip measurement was performed on asphalt roads and on two ice tracks using the continuous roadgrip apparatus RT3 Curve. The aim was to find suitable driving patterns for measurements on public roads and test tracks to ensure the repeatability of roadgrip measurements. During the evaluation, it was concluded that in order to achieve a reliable roadgrip value, regardless of road conditions, the radius of curvature should not be less than 20 m. The velocity dependency of the RT3 Curve is different for the two road conditions, with the measurements on ice being much more sensitive to velocity changes than the measurements on the dry asphalt.

An investigation of a NIR camera system for road surface classification has been conducted for several road conditions. The surfaces were illuminated with three wavelengths, 980 nm, 1310 nm and 1550 nm and a halogen lamp, to simulate a real environment application with surrounding light. A measuring scheme to deal with surrounding light has been implemented enabling road condition classification from NIR images in a real environment. The retrieved camera images have been analyzed and an RGB representation of the different surfaces has been created to classify the different road conditions. The investigation shows that it is possible to distinguish between dry, moist, wet, frosty, icy and snowy road surfaces using a NIR camera system in a disturbed environment.

Three different configurations utilizing polarized short-wave infrared light to classify winter road conditions have been investigated. In the first configuration, polarized broadband light was detected in the specular and backward directions, and the quotient between the detected intensities was used as the classification parameter. Best results were obtained for the SS-configuration. This sensor was shown to be able to distinguish between the smooth road conditions of water and ice from the diffuse road conditions of snow and dry asphalt with a probability of wrong classification as low as 7%. The second sensor configuration was a pure backward architecture utilizing polarized light with two distinct wavelengths. This configuration was shown to be effective for the important problem of distinguishing water from ice with a probability of wrong classification of only 1.5%. The third configuration was a combination of the two previous ones. This combined sensor utilizing bispectral illumination and bidirectional detection resulted in a probability of wrong classification as low as 2% among all four surfaces.

By measuring the spectral reflection from the four different road conditions dry, wet, icy, and snowy asphalt, a method of classification for the different surfaces -- using two and three wavelengths -- is developed. The method is tested against measurements to ascertain the probability of wrong classification between the surfaces. From the angular spectral response, the fact that asphalt and snow are diffuse reflectors and water and ice are reflective are confirmed.

An investigation of different road conditions has been conducted using a short-wave infrared (SWIR) light online sensor to examine the possibility of estimating road condition parameters such as porosity, depth and roughness. These parameters are essential for non-contact road friction estimation. The investigation show that it is possible to detect changes of depths of water and ice as well as classify different types of ice, by utilising polarised short-wave infrared (SWIR) light and a modified Hapke directional reflectance model

Measuring the road condition in front of a vehicle could prevent accidents and make technologies like electronic stability control (ESP) more efficient. By making three investigations of the classifications of the four road conditions dry asphalt, asphalt covered with water, ice and snow the possibility of a preview sensor is exploited. By measuring the reflectance from the different surfaces with a halogen light and an actual sensor (Road eye) in a laboratory surroundings the advantage and disadvantage are revealed. The sensor is also mounted in a Volvo truck for real-life condition measurements.

Water-contaminated grease samples are investigated with attenuation spectra in the visible and near-infrared (NIR) regions in this article. The purpose of this investigation was to identify a model with optical attenuation spectra such that the water content of grease samples could be characterized with a simple measurement setup using common methodology from the field of instrumental chemistry. The ratio between two chosen wavelengths of light appears to approximate the water content of grease samples with an acceptable coefficient of determination using a methodology to show what can potentially be done to develop condition monitoring tools. To illustrate the outlined method, a prestudy of grease aging and oxidation levels is also investigated to show that other variables do not significantly change the measurement.

We describe experiments and numerical modeling of snow surface contamination on two simplified automotive bluff bodies: The Ahmed body and a wedge. The purpose was twofold: 1) To obtain well defined experimental results of snow contamination on simple geometries; 2) To propose a numerical modeling approach for snow contamination. The experiments were performed in a climatic wind tunnel using a snow cannon at −15 °C and the results show that the snow accumulation depends on the aerodynamics of the studied bluff bodies. Snow accumulates on surfaces in proximity to the aerodynamic wakes of the bodies and characteristic snow patterns are obtained on side surfaces. The numerical modeling approach consisted of an aerodynamic setup coupled with Lagrangian particle tracking. Particles were determined to adhere or rebound depending on an adhesion model combined with a resuspension criterion. The adhesion model was based on adhesive-elastic contact theory and the resuspension criterion is derived from the balance between the aerodynamic forces acting on a particle and the critical force for onset of resuspension. The results show that the numerical method can predict certain characteristic snow patterns obtained from the experiments and we also highlight deviations obtained between experimental and simulation results. The simulation results show that the snow accumulation patterns on a bluff body will depend on the smallest ice particles in a snow sample which implies that samples with larger ice particle (for example natural snow) could produce different snow patterns than the fine machine-made snow used in this study.

The angle of repose is a measure reflecting the internal friction and cohesion properties of a granular material. In this paper, we present an experimental setup and measurements for the angle of repose of snow for seven different snow samples over a large range of temperatures. The results show that the angle of repose is dependent on the fall height, the temperature, and the grain size of the snow. These variables are quantified, and their interdependencies are separately studied. With increased snow temperature, the angle of repose increases, and this can be explained by the presence of a liquid layer on ice that can be thermodynamically stable at temperatures below the melting point of water. With decreasing grain size the angle of repose also increases which is expected since the cohesive energy decreases more slowly than the grain mass. For increasing fall height, the snow grains generally accelerate to larger collisional velocities, yielding a smaller angle of repose. In general, the dimensionless cohesion number was found to largely reflect the dependencies of the variables and is therefore useful for understanding what affects the angle of repose. The results demonstrate that the drag force and collision dynamics of ice grains are important for understanding how snow accumulates on a surface, for example if one desires predicting snow accretion by simulating a dispersed cloud of snow.

This paper presents experimental values for the coefficient of restitution (en) for millimeter-sized ice particles colliding with massive walls at different temperatures. Three different wall materials are tested: hardened glass, ice and Acrylonitrile butadiene styrene (ABS) polymer. The results show a high sensitivity to impact velocity Vi, where en decreases rapidly with increasing Vi. The results also show a decrease in en with increasing temperature T. A novel model that predicts en based on the assumption of collisional melting and viscous damping caused by an increased premelted liquid-layer, is proposed. The model predicts both the velocity and the temperature trends seen in the experiments. The difference obtained in experiments between wall materials is also captured by the new model. A generalized regime map for ice particle collisions is proposed to combine the new model with previous work.

Angularly resolved bidirectional reflectance measurements were modelled by ap- proximating a first order Legendre expanded phase function to retrieve single scattering properties of snow. The measurements from 10 different snow types with known density and specific surface area (SSA) were investigated. A near infrared (NIR) spectrometer was used to measure reflected light above the snow surface over the hemisphere in the wavelength region 900 nm to 1650 nm. A solver based on discrete ordinate radiative transfer (DISORT) model was used to retrieve the estimated Legendre coefficients of the phase function and a cor- relation between the coefficients and physical properties of different snow types is investigated. Results of this study suggest that the first two coefficients of the first order Legendre phase function provide sufficient information about the physical properties of snow where the latter captures the anisotropic behaviour of snow and the former provides a relative estimate of the single scattering albedo of snow. The coefficients of the first order phase function were com- pared with the experimental data and observed that both the coefficients are in good agreement with the experimental data. These findings suggest that our approach can be applied as a qualitative tool to investigate physical properties of snow and also to classify different snow types.

The present study focuses on three-dimensional (3D) microstructure analysis of dry natural snow during compaction. An X-ray computed microtomography (micro-CT) system was used to record a total of 1601 projections of a snow volume. Experiments were performed in-situ at four load states as 0 MPa, 0.3 MPa, 0.6 MPa and 0.8 MPa, to investigate the effect of compaction on structural features of snow grains. The micro-CT system produces high resolution images (4.3 μm voxel) in 6 hours of scanning time. The micro-CT images of the investigated snow volume illustrate that grain shapes are mostly dominated by needles, capped columns and dendrites. It was found that a significant number of grains appeared to have a deep hollow core irrespective of the grain shape. Digital volume correlation (DVC) was applied to investigate displacement and strain fields in the snow volume due to the compaction. Results from the DVC analysis show that grains close to the moving punch experience most of the displacement. The reconstructed snow volume is segmented into several cylinders via horizontal cross-sectioning, to evaluate the vertical heterogeneity of porosity distribution of the snow volume. It was observed that the porosity (for the whole volume) in principle decreases as the level of compaction increases. A distinct vertical heterogeneity is observed in porosity distribution in response to compaction. The observations from this initial study may be useful to understand the snow microstructure under applied stress.

The purpose of this paper is to estimate the scattering phase function of snow from angularly resolved measurements of light intensity in the plane of incidence. A solver is implemented that solves the scattering function for a semi-infinite geometry based on the radiative transfer equation (RTE). Two types of phase functions are considered. The first type is the general phase function based on a low-order series expansion of Legendre polynomials and the other type is the Henyey-Greenstein (HG) phase function. The measurements were performed at a wavelength of 1310 nm and six different snow samples were analysed. It was found that a first order expansion provides sufficient approximation to the measurements. The fit from the first order phase function outperforms that of the HG phase function in terms of accuracy, ease of implementation and computation time. Furthermore, a correlation between the magnitude of the first order component and the age of the snow was found. We believe that these findings may complement present non-contact detection techniques used to determine snow properties.

Since Winter Road Maintenance (WRM) is an important activity in Nordic countries, accurate intelligent cost-effective WRM can create precise advance plans for developing decision support systems to improve traffic safety on the roads, while reducing cost and negative environmental impacts. Lack of comprehensive knowledge and inaccurate WRM information would lead to a certain loss of WRM budget, safety reduction, and irreparable environmental damage. This study proposes an intelligent methodology that uses data envelopment analysis and machine learning techniques. In the proposed methodology, WRM efficiency is calculated by data envelopment analysis for different decision-making units (roads), and inefficient units need to be considered for further assessments. Therefore, road surface temperature is predicted by means of machine learning methods, in order to achieve efficient and effective WRM on the roads during winter in cold regions. In total, four different methods have been used to predict road surface temperature on an inefficient road. One of these is linear regression, which is a classical statistical regression technique (ordinary least square regression); the other three methods are machine-learning techniques, including support vector regression, multilayer perceptron artificial neural network, and random forest regression. Graphical and numerical results indicate that support vector regression is the most accurate method.

One of the main challenges in developing efficient and effective winter road maintenance is to design an accurate prediction model for the road surface friction coefficient. A reliable and accurate prediction model of road surface friction coefficient can help decision support systems to significantly increase traffic safety, while saving time and cost. High dynamicity in weather and road surface conditions can lead to the presence of uncertainties in historical data extracted by sensors. To overcome this issue, this study uses an adaptive neuro-fuzzy inference system that can appropriately address uncertainty using fuzzy logic neural networks. To investigate the ability of the proposed model to predict the road surface friction coefficient, real data were measured at equal time intervals using optical sensors and road-mounted sensors. Then, the most critical features were selected based on the Pearson correlation coefficient, and the dataset was split into two independent training and test datasets. Next, the input variables were fuzzified by generating a fuzzy inference system using the fuzzy c-means clustering method. After training the model, a testing set was used to validate the trained model. The model was evaluated by means of graphical and numerical metrics. The results show that the constructed adaptive neuro-fuzzy model has an excellent ability to learn and accurately predict the road surface friction coefficient.

The prediction of efficiency scores for winter road maintenance (WRM) is a challenging and serious issue in countries with cold climates. While effective and efficient WRM is a key contributor to maximizing road transportation safety and minimizing costs and environmental impacts, it has not yet been included in intelligent prediction methods. Therefore, this study aims to design a WRM efficiency classification prediction model that combines data envelopment analysis and machine learning techniques to improve decision support systems for decision-making units. The proposed methodology consists of six stages and starts with road selection. Real data are obtained by observing road conditions in equal time intervals via road weather information systems, optical sensors, and road-mounted sensors. Then, data preprocessing is performed, and efficiency scores are calculated with the data envelopment analysis method to classify the decision-making units into efficient and inefficient classes. Next, the WRM efficiency classes are considered targets for machine learning classification algorithms, and the dataset is split into training and test datasets. A slightly imbalanced binary classification case is encountered since the distributions of inefficient and efficient classes in the training dataset are unequal, with a low ratio between classes. The proposed methodology includes a comparison of different machine learning classification techniques. The graphical and numerical results indicate that the combination of a support vector machine and genetic algorithm yields the best generalization performance. The results include analyzing the variables that affect the WRM and using efficiency classes to drive future insights to improve the process of decision-making.

Road surface temperature (RST) plays an essential role in analyzing road surface conditions during winter in countries with adverse winter climates. A reduction in RST can have a negative impact on road safety due to decreasing vehicle grip on the road surface. Therefore, decision makers need to monitor low surface temperatures and plan for winter road maintenance. However, RST sensors can fail for different reasons, such as power outages. RST sensor failure will lead to lack of information about the road surface, which can be problematic, especially for critical road segments. Hence, the novelty of this study is to use a deep learning algorithm to predict RSTs in road segments if a sensor fails at a station using historical data from two other road stations. The mean absolute error in the proposed model is 0.453 and the model explains 98.6% of observations. In addition, since the adjustment of deep learning parameters (e.g., hidden layers, optimizer, activation function, etc.) is associated with epistemic uncertainty, a semiquantitative approach is developed for uncertainty assessment. With this approach, the most important and uncertain parameters in RST prediction models can be identified. The results have shown that the optimizer is the most uncertain and important parameter.

The decade of big data has emerged in recent years, which has led to entering the era of intelligent transportation. One of the main challenges to deploying intelligent transportation is dealing with winter roads in cold climate countries. Different operations can be used to protect the road from ice and snow, such as spreading chemicals (here salt) on the road surface. Using salt for de-icing and anti-icing increases road safety. However, the excess use of salt must be avoided since it is not cost-efficient and has negative impacts on the environment. Therefore, the accurate and timely prediction of salt quantity for winter road maintenance helps decision support systems to achieve effective and efficient winter road maintenance. Thus, this paper performs exploratory data analysis to determine the relationships among variables to find the best prediction model for this problem. Due to the stochastic nature of variables regarding weather and roads, a deep neural network/deep learning is selected to predict the amount of salt on the wheel track, using historical data measured by sensors and road weather stations. The results show that the proposed model performs perfectly to learn and predict the amount of salt on the wheel track, based on different metrics, including the loss function, scatter plot, mean absolute error, and explained variance.

There is a need for an automated road status classification system considering the vast number of weather-related accidents that occur every winter. Previous research has shown that it is possible to detect hazardous road conditions, including, for example, icy pavements, using single point infrared illumination and infrared detectors. In this paper, we extend this research into camera surveillance of a road section allowing for classification of area segments of weather-related road surface conditions such as wet, snow covered, or icy. Infrared images have been obtained using an infrared camera equipped with a set of optical wavelength filters. The images have primarily been used to develop multivariate data models and also for the classification of road conditions in each pixel. This system is a vast improvement on existing single spot road status classification systems. The resulting imaging system can reliably distinguish between dry, wet, icy, or snow covered sections on road surfaces.

The real friction coefficients between the rails and the wheels on a 360. t and 10,800. kW IORE locomotive were measured using the locomotive[U+05F3]s in-built traction force measurement system. The locomotive consisted of two pair-connected locomotives had a CoCo+CoCo bogie configuration, and hauled a fully loaded set of 68 ore wagons (120. t/wagon). The measurements were performed both on rails in a dry condition and on rails lubricated with a water-based top-of-rail (ToR) friction modifier on the Iron Ore Line between the cities of Kiruna and Narvik in Northern Sweden and Norway, respectively. Since full-scale measurements like these are costly, the friction coefficients were also measured at the same time and place using a conventional hand-operated tribometer, with and without the ToR friction modifier. The most important results are that the real friction coefficient is definitely not constant and is surprisingly low (0.10-0.25) when the ToR friction modifier is used, and that it is also significantly dependent on the amount of ToR friction modifier. A large amount will reduce the friction coefficient. Furthermore, it is concluded that the real friction coefficients are in general lower than the friction coefficients measured with the hand-operated tribometer. A final remark is thus that the use of a water-based ToR friction modifier can give excessively low friction, which can result in unacceptably long braking distances.

The appearance of ground frost is of vital importance in construction and maintenance of roads in cold climates. Frost often causes ground heave and subsequent road damage, which must be taken into account in designing the road structure. Frost depth, pavement temperature, and freezing/thawing cycles are also important for estimating the frequency of road maintenance and treatment. Various analytical, numerical, and empirical models have been developed to estimate the surface temperature of the pavement and to model the heat flow in the underlying layers. The pavement surface experiences a variety of intricate nonlinear heat transfer mechanisms during winter, making it challenging to accurately model the surface boundary. Dynamic variation of parameters such as cloud cover and traffic density during the modeling period introduces additional complexity. To address this challenge, we have established an experimental setup in Luleå, Sweden, to measure pavement profile temperatures during the winter season. Additionally, we have developed a Finite Difference Model that utilizes local weather data including dynamic cloud cover, and which also takes traffic into account. The experimental and simulation findings demonstrate how the impact of surface temperature fluctuations diminishes and, more or less, vanishes for depths more than 55 [cm] below the pavement surface. The Finite Difference Model presented in this study exhibits the ability to forecast the pavement profile temperatures, including the surface temperature based on weather conditions, with acceptable precision for at least 3 days. As a consequence, a reasonable assessment of pavement layer conditions appears feasible based on local weather conditions, and the model can serve as a useful tool for planning road maintenance and construction in cold regions.

Pressure-sinkage tests for determining vehicle sinkage on soft soils can be done using a bevameter. In this study, pressure-sinkage tests were performed on snow, which, like soil, is a granular material. However, unlike soil, snow layers are inhomogeneous with varying properties. For tracked vehicles, the shape of the track print is rectangular, which is why rectangular plates are often used for pressure-sinkage tests. The aim of this study was to see if smaller circular plate or smaller rectangular plates can be used instead of larger rectangular plates, and to understand the possible limitations of using small plates. Radius for the circular plate was chosen to be equal to the width of the rectangular plate. Three measuring sessions were performed at different locations during different snow conditions using circular pressure plates and rectangular pressure plates of different aspect ratios. The results show that smaller rectangular plates can be used if the width of the plates remains the same, or circular plates can be used if the radius of the circular plate is equal to the width of the rectangular plate. Limitation comes with increasing pressure, which occurs more quickly with larger-area plates, as larger plates sense solid ground more rapidly than smaller plates. To avoid this, snowpack thickness should be a minimum of five times thicker than maximum sinkage.

Mechanical properties of snow related to snowmobiles or similar lightweight tracked vehicles aren't widely researched today and it is difficult to find data. One challenge is that snow properties constantly are changing due to aging, climate conditions and location. Also the measuring procedure is difficult since aged snow often contains layers with various densities and hardness. Soil is to some extent similar to snow, in the context that both are granular materials. The bevameter is a popular device for measuring soil properties, however this device needs to be scaled in order to meet criteria of target for research, i.e. in this case snowmobiles. In this paper a new type of portable bevameter is presented, which is designed and built for measuring snow properties in the field. Results from initial tests are also presented. The aim with the bevameter is to measure snow properties which can be used to simulate the interaction between a snowmobile and soft snow. The designed bevameter can be towed with one snowmobile to the field to execute measurements. One full set of test results is introduced and parameters for simulations are extracted from the result data. The parameters from the data were usable but the quality of the measurements can be improved. One problem with the data collected was noise, which was caused by the interaction between the mechanical parts and the low mass of the bevameter. Furthermore, the usability can be improved by reducing cables which can be hard and fragile during cold weather and by replacing the laser distance-sensor with a string wire potentiometer which isn't sensitive to snow dropping in the measurement area. With some improvements the constructed bevameter is a very useful tool which can be used for field measurements to determine snow properties for snowmobile-size vehicle simulations.

Purpose: For the expected increase in the capacity of existing transportation systems and efficient energy utilisation, smart maintenance solutions that are supported by online and integrated condition monitoring systems are required. Industrial Internet is one of the smart maintenance solutions, which enables real-time acquisition and analysis of asset condition by linking intelligent devices with different stakeholdersￜ applications and databases. This paper presents some aspects of Industrial Internet application as required for integrating weather information and floating road condition data from vehicle mounted sensors to enhance effective and efficient winter maintenance.

Design/methodology/approach: The concept of real-time road condition assessment using in-vehicle sensors is demonstrated in a case study of a 3.5 km road section located in northern Sweden. The main floating data sources were acceleration and position sensors from a smartphone positioned on the dash board of a truck. Features extracted from the acceleration signal were two road roughness estimations. To extract targeted information and knowledge, the floating data were further processed to produce time series data of the road condition using Kalman filtering. The time series data were thereafter combined with weather data to assess the condition of the road.

Findings: In the case study, examples of visualisation and analytics to support winter maintenance planning, execution, and resource allocation were presented. Reasonable correlation was shown between estimated road roughness and annual road survey data to validate and prove the presented results wider applicability.

Originality/value: The paper describes a concept of floating data for an industrial internet application for efficient road maintenance. The resulting improvement in winter maintenance will promote dependable, safe and sustainable transportation of goods and people, especially in northern Nordic region with harsh and sometimes unpredictable weather conditions.

Low adhesion between railway wheel and rail, which is usually induced by contaminants such as water, oil, leaves etc., affects railway operation in terms of performance and safety. This study uses an optical sensor to identify different surface layers which cause low adhesion. A laboratory set up and field tests under various conditions were subject to the surface layer identification by the optical sensor. In addition, the friction coefficient was measured on the same surface layers. The results show that the sensor can distinguish between different surface layers. This information further linked to the levels of the friction coefficient, which can be used in the prediction system for the railway operator.