Design and development of software stack of an autonomous vehicle using robot operating system

In recent research activities, autonomous vehicles and self-driving technology have gained lot of attention among scientists. The idea of autonomous vehicles can be anticipated in the 1920s when the design of the first radio-controlled vehicles was in progress. Autonomous vehicles are going to be the trend of the future in this modern era of automation and technology. In this paper various autonomous driving aspects, highlighting the software stack and hardware components are discussed. The software architecture covers mainly robot operating system (ROS), machine learning (ML), deep learning (DL), and OpenCV frameworks, along with the calibration of sensors and cameras. The paper also discussed about simultaneous localization and mapping (SLAM) based-path tracking, computer vision-based controller, and intelligent object avoidance. Further, point cloud, ground, radius, and raycast filters was implemented to distinguish between the real-time objects, ground, and its own parts or obstacle shadows. The paper highlights the overall hardware modules responsible for controlling the car.(c) 2022 Elsevier B.V. All rights reserved.