Investigating the design patterns of sensor sets in self-driving cars

Self-driving cars have been under intensive development in the last few years. These vehicles combine a variety of sensors, such as radar, lidar, sonar, and camera, to understand their surroundings. Each vehicle uses different sensors and installation locations. Although there is a great deal of variation in the set of sensors and their installation locations, some repetitive patterns can be seen among self-driving cars. In this paper, these patterns are examined and their different approaches are explained. First, the patterns used in some of the existing vehicles are presented. Then, general considerations in the design of sensor patterns are expressed. Next, the advantages and disadvantages of each pattern are examined. Finally, the results of this study are summarized.