Autonomous vehicular overtaking maneuver: A survey and taxonomy

Abstract

Autonomous vehicles (AVs) are the next-generation driver-less vehicular entities with advanced technologies. Overtaking is an important and challenging maneuver that needs to be frequently performed in diverse environments during the run. The maneuvering of AVs, in present scenarios, improves driving precision, fuel efficiency, and travel time and minimizes road accidents caused by human error, waiting time, carbon emission, etc. Despite all these benefits, AVs face several critical issues, such as traffic safety, market issues, environmental aspects, technical compatibility, market introduction, etc. In traffic safety, the overtaking scenario is considered one of the most complex driving maneuvers out of several driving maneuvers, such as lane changing, lane following, tailgating, and many more. Despite this, to the best of authors' knowledge, an extensive study on autonomous vehicular overtaking maneuvers still needs to be explored. Therefore, this survey aims to visualize the current methods used to enhance traffic safety in an intelligent transport system by handling the most complex driving maneuver in autonomous driving (overtaking). In this survey, we present the taxonomy, simulators, and methods used for AVs overtaking maneuver, and the state-of-the-art methods are further categorized under theoretical and AI-based techniques. Furthermore, the theoretical and AI-based techniques are classified based on the applications of the respective technique in overtaking maneuvers. While designing this survey, several theoretical and practical studies are taken into consideration. As the outcomes of this study, several research gaps, challenges, future research directions, and open research problems in overtaking maneuvers are identified. The outcomes of this study would be helpful for the researchers to carry out their research in the domain of AVs overtaking maneuvers.