Architectural Comparison of ROS1 and ROS2 in the Context of Processing LiDAR Point Cloud Data

Authors: Korshunova E. V., Syrkin I. S., Sadovets V. Yu.

Annotation: This paper provides a comparative analysis of the architectural features of ROS1 and ROS2 as applied to the task of pro-cessing LiDAR point clouds. Considering the mining industry's requirements for reliability, scalability, and fault tolerance, the key components of both ROS versions, their architectural dif-ferences, the limitations of ROS1, and the advantages of ROS2 are discussed. Special attention is given to the implementation of the LiDAR data preprocessing pipeline in ROS2, integration with SLAM and object detection modules, as well as DDS configuration and QoS settings. The article also discusses dif-ficul-ties encountered when migrating from ROS1 to ROS2 and offers recommendations for overcoming them, including the use of bridges for joint operation of nodes. The presented results can be useful in designing reliableand efficient control systems for unmanned vehicles operating in open-pit mines and other complex industrial environments.

Keywords: rviz, pointcloud_preprocessor, slam, point cloud, lidar, ros