Pavement slipperiness detection using wheel speed and acceleration sensor data

Abstract

Faced with high rates of traffic accidents on slippery roads, it is recommended that road managers promptly identify the slippery spots, remove the slipperiness and inform drivers of the location of the dangerous spots on the route ahead, to allow them to prepare for the slipperiness. In this paper, the wheel slip-based and wheel acceleration-based approaches are suggested for detecting road slipperiness using sensor data from a digital tachograph (DTG), which is mandatory in commercial vehicles. Support vector machine algorithms were employed to categorize slippery and non-slippery states. The performances were outstanding (with accuracy more than 98%) when evaluated using experimental data, collected on a road weather-proving ground. Considering a large number of commercial vehicles equipped with DTG devices which are connected via cellular communications, the suggested methods could be practically applied to real-world DTG-based fleet management systems.