Mapping with ROS Noetic on Ubuntu 20.04

Overview

The AWS DeepRacer Mapping sample project provides instructions for using the AWS DeepRacer and Intel RealSense™ D435, along with other open-source tools, to build a map of your surroundings using SLAM (Simultaneous Localization and Mapping).

License

The source code is released under Apache 2.0.

Setup

Follow these instructions to set up the AWS DeepRacer Mapping sample project.

Prerequisites

The AWS DeepRacer device comes with all the prerequisite packages and libraries installed for running the AWS DeepRacer core application. For more information about the preinstalled set of packages and libraries on the AWS DeepRacer, and about installing the required build systems, see Getting started with AWS DeepRacer OpenSource. The Mapping sample project requires the AWS DeepRacer application to be installed on the device, as it leverages the manual drive feature to control the car manually.

Intel RealSense™ D435 / D435i

Connect and mount the Intel RealSense D435 or Intel RealSense D435i to the AWS DeepRacer device. The Intel RealSense cameras provide a depth sensor and other required capabilities to collect information from your surroundings.

ROS Noetic on the AWS DeepRacer device

Open a terminal and install ROS Noetic as a root user by following the steps on the Ubuntu install of ROS Noetic page.

Download and install other components

Open a terminal and run the following commands as the root user on the AWS DeepRacer device:

• Set up the ROS Noetic environment: After installing ROS Noetic on your AWS DeepRacer, source the setup script for ROS Noetic:

        source /opt/ros/noetic/setup.bash
  

• Install the software packages required to create the map:

  • realsense2_camera:

      export ROS_VER=noetic
      sudo apt-get install ros-$ROS_VER-realsense2-camera
      sudo apt-get install ros-$ROS_VER-realsense2-description
    

  • imu_filter_madgwick:

      sudo apt-get install ros-$ROS_VER-imu-filter-madgwick
    

  • rtabmap_ros:

      sudo apt-get install ros-$ROS_VER-rtabmap-ros
    

  • robot_localization:

      sudo apt-get install ros-$ROS_VER-robot-localization
    

Usage

Follow these steps to use the AWS DeepRacer Mapping sample project.

• Open a terminal and run the following commands as the root user on the AWS DeepRacer device:

    source /opt/ros/noetic/setup.bash
  

• Initiate the open-source tracking launch script provided by Intel RealSense:

    roslaunch realsense2_camera opensource_tracking.launch
  

• Open another terminal and launch rviz by running the following command as the root user:

    source /opt/ros/noetic/setup.bash
    rosrun rviz rviz
  

• Personalize, collect, and visualize the point cloud data by following these steps.

• Move the AWS DeepRacer car slowly around the room in manual mode using the device console.

Sample demo:

The following example shows using rviz when mapping the room by navigating the AWS DeepRacer car:

Resources

• Getting started with AWS DeepRacer OpenSource
• Installing ROS Noetic