National Concrete Pavement Technology Center

Project description

Measuring roadway friction is crucial for roadway safety, particularly on wet surfaces and challenging road geometries. Roadway friction, influenced by pavement design, aggregate type, traffic loading, surface treatment, and weather, fluctuates over time. Traditionally, state agencies perform costly, periodic friction measurements using specialized devices and vehicles. This project introduces an advanced road friction screening system using telematics data from both regular and automated vehicles (RVs and AVs), enabling comprehensive network-level friction analysis. The system utilizes Physics-Enhanced Residual Learning (PERL) for AV control to maintain optimal slip ratios and peak Tire-Road Friction Coefficient (TRFC) values, ensuring accurate friction measurements without causing excessive sliding or tire wear. It leverages cooperative perception from connected vehicles to improve accuracy and extend coverage by mitigating sensor errors. Additionally, the system employs smart routing to optimize data collection routes for connected AVs, enhancing coverage and the efficiency of the screening process. The system's potential to improve road safety, cooperative driving automation (CDA) applications, and infrastructure management will be demonstrated through field tests with real-world scenarios in UW-Madison's Level 3 testbed and possibly later at Mcity.