Road Surface Characterization using Crowdsourcing Vehicles

The contact between vehicle's tires and the road surface is one of the crucial elements for driving safety and efficiency. The contact's quality depends on several factors, some of them relate to the vehicle itself and some relate to the surface conditions. Generally, these parameters change gradually with time and need to be monitored. Therefore, efforts are made to map road surfaces and analyze the collected data.
A traditional way of detecting and mapping road surface is by using dedicated survey vehicles. In recent years, smart vehicles are becoming more and more ubiquitous. Such vehicles are generally equipped with a multitude of sensors, from wheel speed to camera and LiDAR in addition to providing internet connectivity. Thus, the use of crowd-sourcing techniques for mapping road and vehicle characteristics is becoming more practical. However, on-board vehicle sensors are normally inaccurate and are influenced by vehicle and weather conditions.
MobiWize Solutions has developed a novel approach for road surface characteristics mapping, using crowd sourcing and sharing and demonstrated correlation with survey data provided by the Israel National Transport Infrastructure Company. This paper describes the challenges and limitations of existing characterization solutions, methods to overcome them and finally provides field tests results.