Road Surface Friction

Accurately estimating road surface friction is crucial for enhancing vehicle safety and control, particularly in challenging winter conditions or for autonomous driving systems. Current research heavily utilizes computer vision, employing deep learning architectures like convolutional neural networks and transformers to predict friction coefficients from images captured by roadside or onboard cameras, often incorporating uncertainty estimation for improved reliability. These advancements, along with data-driven approaches using sensor data and machine learning regression models, aim to improve braking systems and provide real-time road condition information for safer and more efficient transportation.