Pedestrian and Non-motor Vehicle Detection Systems with Lidar for Autonomous Vehicles

The purpose of this paper is to examine the use of Light Detection And Ranging (LiDAR) in obstacle detection systems in autonomous vehicles (AVs). Particular attention will be given to methods being developed to detect nonmotorized bodies such as pedestrians and cyclists due to unique variables they present compared to other vehicles. Due to the increasing societal and professional relevance of AV development, we examine LiDAR’s current role in the development of pedestrian detection systems in autonomous vehicles as well as its implications for sustainable development. We begin our paper with an overview of AV development and the role of LiDAR within it. We then survey currently-proposed LiDAR-based detection methods for nonmotorized bodies and critically examine the potential technical problems associated within. This is followed by an examination of machine-learning techniques being implemented to decrease errors in detection. Waymo, formerly Google’s self-driving car, is used as a case study. We conclude with a discussion of current technical, social, and ethical problems associated with AVs and the possible solutions to reach a qualified position on LiDAR’s role in future AV development.