CrowdParking: Crowdsourcing based parking navigation in autonomous driving era

Abstract

Finding a free road side parking in urban area is considered as one of the most challenging driving tasks, especially for the autonomous vehicles with limited sight (e.g. short range sensing) and brain (compared with human beings). To assist autonomous vehicle parking in urban area, we propose a novel parking scheme CrowdParking, which applies crowdsourcing and vehicular fog computing to collect parking information from vehicles, locate free parking spaces from crowdsourced data. We also explore the variation of parking availability from a real world data set and find that the availability of specific parking lot has certain relationship with the traffic condition of nearby roads. Based on the observations, we propose the vision of estimating the parking availability with taking into account the traffic condition in neighborhood.