Road transport safety and comfort have gradually become the focus of social attention with the rapid development of superhighways and intelligent driving technology. To identify the research status and hotspots in skid resistance performance of asphalt pavement, this study reviews the tire‒road contact mechanism, texture deterioration behavior, and skid resistance mapping relationship, multi-factor coupled skid resistance early warning model, skid-resistant durable pavement design, and related research. The analysis shows that there are still some problems in the current research, such as an insufficient mapping relationship between mixture and texture, nonuniform testing methods, and an idealized predictive model. The Persson’s hysteretic friction model provides the most comprehensive calculation for skid resistance mechanics at present, and the Permanent International Association of Road Congresses (PIARC) model has the strongest correlation and normalization with various test methods. The skid resistance early warning model based on the analysis of temporal and spatial texture deterioration is currently a popular research topic and will provide theoretical guidance for the design of functional, durable pavement.

Current methods for assessing pavement skid resistance are based on spot or line sampling, neglecting the lateral skidding risk of vehicles derived from the uneven distribution of pavement friction coefficients. Through mechanical analysis, this study illustrates that vehicles are susceptible to lateral instability when the road surface exhibits unequal friction coefficients between the left and right wheel tracks. On the basis of these findings, vehicle dynamics simulations are conducted to evaluate longitudinal and lateral braking distances under varying speeds and friction coefficient distributions. When the friction coefficient is less than 0.5, the risk is dominated by the longitudinal braking distance. Conversely, when there is a significant disparity in friction coefficients between the left and right wheel tracks (exceeding 0.2), the risk is predominantly associated with lateral skidding. A sensitivity analysis further examined the combined effects of friction disparities and driving speed, revealing that when the speed exceeds 80 km/h, the lateral skidding risk induced by uneven friction becomes the dominant factor over the longitudinal braking risk. A skidding risk assessment method is then proposed, incorporating simulation results and braking distance thresholds. Furthermore, a comprehensive evaluation framework is established, encompassing sampling strategies, paired friction coefficient analysis for left and right wheel tracks, and risk quantification. The key contribution of this study lies in highlighting the critical yet often neglected impact of lateral friction coefficient variation on vehicle skid safety. By simulating its risk implications, this research proposes a novel overall evaluation framework for pavement skid resistance, leveraging field-collected data. The proposed approach expands the scope of traditional skid resistance assessment, offering a more holistic perspective for improving road safety.

Understanding the tire‒road friction system is fundamental for evaluating the skid resistance of asphalt pavements. Literature analysis reveals that the trajectory of tire–road friction research aligns with the evolution of scientific research paradigms: experimental science, theoretical science, computational science, and data science. Research in this field can be categorized into three scales: the rubber‒pavement scale, the tire‒road scale, and the vehicle scale. Experimental observations have yielded numerous patterns and empirical models, which serve as the foundation of this research field. Although numerical measurement devices have been used for decades, the reproducibility and comparability of the results require further improvement. Tire‒road friction theory and simulations have been well developed across these three scales, but these scales remain largely independent and unconnected. With the advancement of sensing technology, texture features have been widely exploited and used as inputs for various machine learning models to estimate pavement skid resistance. However, these models are limited in their ability to integrate friction mechanisms, resulting in relatively low interpretability. In summary, the synergistic development of the four research paradigms can promote and advance the understanding and application of tire‒road friction mechanisms. This review concludes with a discussion of current challenges and future trends, drawing implications for further research in this field.