This paper evaluates the ability of some state-of-the-art Machine Learning models, namely SVM (support vector machines), DT (decision tree) and MLR (multiple linear regression), to predict pavement skid resistance. The study encompasses both regression and classification tasks. In the regression task, the aim is to predict the coefficient of friction values, while the classification task seeks to identify three classes of skid resistance: good, intermediate and bad. The dataset used in this work was gathered through an extensive test campaign that involved a fifth-wheel device to measure the coefficient of friction at different slip ratios on different road surfaces, vehicle speeds, tire tread depths and water depths. It was found that the RBF-SVM model, due to its ability to capture non-linear relationships between the features and the target for a relatively small dataset, is the most adapted tool compared with, on one side, MLR, linear SVM and DT models for the regression task and, on the other side, linear SVM and DT models for the classification task. The paper also discusses the strengths and weaknesses of the investigated models based on the underlying physical phenomena related to skid resistance.

This paper deals with the modeling of rolling resistance and the analysis of the effect of pavement texture. The Rolling Resistance Model (RRM) is a simplification of the no-slip rate of the Dynamic Friction Model (DFM) based on modeling tire/road contact and is intended to predict the tire/pavement friction at all slip rates. The experimental validation of this approach was performed using a machine simulating tires rolling on road surfaces. The tested pavement surfaces have a wide range of textures from smooth to macro-micro-rough, thus covering all the surfaces likely to be encountered on the roads. A comparison between the experimental rolling resistances and those predicted by the model shows a good correlation, with an R² exceeding 0.8. A good correlation between the MPD (mean profile depth) of the surfaces and the rolling resistance is also shown. It is also noticed that a random distribution and pointed shape of the summits may also be an inconvenience concerning rolling resistance, thus leading to the conclusion that beyond the macrotexture, the positivity of the texture should also be taken into account. A possible simplification of the model by neglecting the damping part in the constitutive model of the rubber is also noted.

The paper points out the limits of current approaches to predict the evolution of skid resistance over time from the road surface texture. Simulation in laboratory of traffic-induced polishing shows that the friction coefficient measured on an asphalt concrete first increases, reaches a maximum then decreases. In the meantime, texture parameters evolve in a scattered way and do not reflect the friction evolution. Comparison of scanning electron photos at different polishing stages reveal cracking in the bitumen layer and the aggregates and fracture at late polishing stages. These findings prove that wear of road surfaces results from more complex processes than just a modification of the surface texture. Considering asphalt concretes as a coated system, with the bitumen deposited as a coating on the substrate composed of aggregates, a more systematic approach is derived to better understand the basic mechanisms. At the micro-scale, similarities in terms of debris formation allows to assume that fatigue wear induces the damages observed in the microstructure. At the macro-scale, this process translates into an evolution of the wear depth, which also highlights the effect of the aggregate roughness in the removal of the bitumen layer. Discussions are finally made on a friction model developed for coated surfaces and its application to predict the evolution of tire/road friction. The proposed study and its first findings show that there is a need to reconsider the texture-friction relationship; this questioning allows to look at an asphalt concrete from a different perspective—considering it as a coated system—and open perspectives for future research.

This paper describes the conception of a high level, compact, scalable, and long autonomy perception and localization system for autonomous driving applications. Our benchmark is composed of a high resolution lidar (128 channels), a stereo global shutter camera, an inertial navigation system, a time server, and an embedded computer. In addition, in order to acquire data and build multi-modal datasets, this system embeds two perception algorithms (RBNN detection, DCNN detection) and one localization algorithm (lidar-based localization) to provide real-time advanced information such as object detection and localization in challenging environments (lack of GPS). In order to train and evaluate the perception algorithms, a dataset is built from 10,000 annotated lidar frames from various drives carried out under different weather conditions and different traffic and population densities. The performances of the three algorithms are competitive with the state-of-the-art. Moreover, the processing time of these algorithms are compatible with real-time autonomous driving applications. By providing directly accurate advanced outputs, this system might significantly facilitate the work of researchers and engineers with respect to planning and control modules. Thus, this study intends to contribute to democratizing access to autonomous vehicle research platforms.

High Friction Surfaces (HFS) are applied to increase friction capacity on critical roadway sections, such as horizontal curves. HFS friction deterioration on these sections is a safety concern. This study deals with characterization of the aggregate loss, one of the main failure mechanisms of HFS, using texture parameters to study its relationship with friction. Tests are conducted on selected HFS spots with different aggregate loss severity levels at the National Center for Asphalt Technology (NCAT) Test Track. Friction tests are performed using a Dynamic Friction Tester (DFT). The surface texture is measured by means of a high-resolution 3D pavement scanning system (0.025 mm vertical resolution). Texture data are processed and analyzed by means of the MountainsMap software. The correlations between the DFT friction coefficient and the texture parameters confirm the impact of change in aggregates’ characteristics (including height, shape, and material volume) on friction. A novel approach to detect the HFS friction coefficient transition based on aggregate loss, inspired by previous works on the tribology of coatings, is proposed. Using the proposed approach, preliminary outcomes show it is possible to observe the rapid friction coefficient transition, similar to observations at NCAT. Perspectives for future research are presented and discussed.

The paper presents an experimental study conducted to evaluate the feasibility of using accelerometers as an indirect means to estimate water depths on road surfaces. It makes use of the vibration of the vehicle’s wheel arch due to water droplets projected by a tire rolling on a wet road surface. A trailer equipped with a wheel and towed by a van was used. The test setups to spread water on the road surface and before the test wheel, measure the water depth and visualize the water spray are described. The test program, conducted on a test track closed to the traffic, includes three surfaces and two speeds. Visualization of water flows by means of high-speed cameras makes it possible to choose a suitable location for the accelerometers. It turns out that signals provided by the accelerometers are affected by the trailer’s movement; a filtering method has been successfully developed to remove noises. Results show a tight relationship between the mean amplitude of accelerometric signals and actual water depths. Discussions are made in terms of effects of the vehicle speed and the road surface texture. Perspectives for using the developed system to improve passenger safety under autonomous driving conditions are presented.