Estimating the Tire–Pavement Grip Potential From Vehicle Vibrations

Abstract

This study was motivated by the desire to provide highway managers/operators with more frequent and spatially dense information about the prevailing friction conditions in their networks. A new data-driven method was outlined for this purpose, wherein the prevailing tire–pavement grip potential is estimated from vehicle vibrations recorded during normal/regular usage of the infrastructure. The method was based on the underlying premise that transverse vehicle accelerations are related to wheel side-force oscillations, and therefore carry information related to the ride surface texture. It involved performing a short-time Fourier transform over vibration signals and analyzing the resulting spectral amplitudes. Two field experiments were carried out to validate the method. The first provided evidence of a statistical link between transverse vehicle vibrations and wheel side-force oscillations. The second tested the statistical link between skid resistance measured over a 26 km highway section and corresponding skid resistance estimations based on vehicle vibration data. Overall, transverse vehicle vibration characteristics were found to hold relevant information about the prevailing tire–pavement grip potential; the two were moderately inter-correlated. The newly proposed estimation method seems promising and potentially useful for pavement management applications, especially when considering the emergence of connected car technologies and the increased availability (and affordability) of in-vehicle Internet of Things devices.