ROS-Embedded LiDAR Perception in CARLA with AGL Visualization

This project aims to build a complete end-to-end LiDAR perception pipeline for Automotive Grade Linux (AGL) using ROS 2 and the CARLA simulator by establishing a Hardware-in-the-Loop (HIL) architecture where CARLA generates LiDAR data on a host system, which is processed through a ROS 2-based perception pipeline and visualized on an AGL-powered Raspberry Pi via a Flutter-based HMI. The system will include a PCL-based LiDAR detection pipeline, real-time visualization interface, and a Yocto layer (meta-agl-lidar) for seamless integration into AGL builds. Deliverables include a ROS 2 perception node, Flutter HMI, RViz configuration, Yocto integration layer, and comprehensive documentation to enable reproducibility and adoption by the AGL community.