About

This repository is for the ECU-Core part of the Autonomous-Driving-System project. The ECU-Core is based on an independent RaspberryPi board and performs following tasks.

• Publish LIDAR scan data as ROS2 topic
• Publish depth camera image data as ROS2 topic
• Receive vehicle informations as ROS2 topic and convert them into CAN communication
• Control the execution of autonomous driving mode

We used YDLIDAR X4 and RealSense D435 depth camera in this project. For more detailed information, refer to the Autonomous-Driving-System project.

Requirements

• Ubuntu 20.04

  Install Ubuntu 20.04 for RaspberryPi using RaspberryPi OS Imager.

• CAN HAT setup

  Follow the instruction of 2-CH CAN FD HAT setup and enable Single SPI Mode.

• ROS2 setup

  Follow the instruction of ROS2 foxy setup.

• LIDAR SDK setup

  Follow the instruction of YDLidar-SDK setup.

• Depth camera SDK setup

  Follow the instruction of RealSense documentation. If you encounter issues while running the software installed on RaspberryPi, refer to the following page.

• OpenCV packages

  pip install opencv-python
  sudo apt install ros-foxy-cv-bridge

• Python packages

  pip install numpy
  pip install transforms3d

Usage

# Execute on the ECU-Core
colcon build
source install/setup.bash
sh can_setup.sh

ros2 run lidar_pub lidar_pub

ros2 run camera_pub camera_pub

ros2 run receiver receiver

ros2 run headunit_start headunit_start

Note

The ROS2 topic communication is machine-to-machine. Make sure that both ECU-Core and local machine are connected to the same WLAN. If the connection is not successful, disable the firewall using the following command.

# Execute on the ECU-Core and local machine
sudo ufw disable

Context of CAN communication

Message	Purpose	Arbitration ID
steering	Control	0x00
throttle	Control	0x01
x position	GPS	0x02
y position	GPS	0x03
orientation	GPS	0x04
headunit start	Autonomous driving	0x05