Skip to content

Latest commit

 

History

History
176 lines (123 loc) · 5.35 KB

README.md

File metadata and controls

176 lines (123 loc) · 5.35 KB
Raw

ORB-SLAM3-Enabled Visual Odometry with Boston Dynamics Spot Robot in Gazebo

Spot Robot Simulation with ORB-SLAM3

This guide provides instructions for setting up a simulation of the Boston Dynamics Spot robot using a ROS-based setup in Gazebo, enhanced with ORB-SLAM3 visual odometry.

Part 1: Boston Dynamics Spot Robot Setup

Prerequisites

  • Ensure you have sudo privileges on your system.
  • ROS must be installed on your system.

Installation Steps

  1. Install Dependencies:

    • Create a catkin workspace if you don't already have one: 
      cd ~/catkin_ws/src
git clone -- https://github.com/ebasatemesgen/ORB-SLAM3-Enabled-Visual-Odometry-.git
    • Install Python ROS dependencies: 
      sudo apt install -y python-rosdep

  2. Install ROS Dependencies:

    cd ~/catkin_ws
rosdep install --from-paths src --ignore-src -r -y

  3. Build the Workspace:

    catkin_make

  4. Clone the Gazebo Models and Worlds Collection Repository:

    git clone https://github.com/leonhartyao/gazebo_models_worlds_collection.git

  5. Update Environment Variables:

    • Update the GAZEBO_MODEL_PATH and GAZEBO_RESOURCE_PATH in your ~/.bashrc file: 
      echo "export GAZEBO_MODEL_PATH=\$GAZEBO_MODEL_PATH:<path to gazebo_models_worlds_collection repo>/models" >> ~/.bashrc
echo "export GAZEBO_RESOURCE_PATH=\$GAZEBO_RESOURCE_PATH:<path to gazebo_models_worlds_collection repo>/worlds" >> ~/.bashrc
source ~/.bashrc
    • Replace <path to gazebo_models_worlds_collection repo> with the actual path.

Usage

  • Run the Spot robot simulation in Gazebo: 
    roslaunch spot_gazebo spot.launch world:=<world_file_name>
  • Replace <world_file_name> with your desired Gazebo world file name.

Customization

  • You can use any world file from the gazebo_models_worlds_collection repository or create custom Gazebo worlds.

Part 2: ORB-SLAM3-ROS

ORB-SLAM3-ROS is a ROS implementation of ORB-SLAM3 V1.0, focusing on the ROS integration. This package uses catkin_make and is tested on Ubuntu 20.04.

1. Prerequisites

  • Eigen3:

    sudo apt install libeigen3-dev

  • Pangolin:

    cd ~
git clone https://github.com/stevenlovegrove/Pangolin.git
cd Pangolin
mkdir build && cd build
cmake ..
make
sudo make install

  • OpenCV:

    • Check your OpenCV version (minimum 3.0 required): 
      python3 -c "import cv2; print(cv2.__version__)"
    • Follow the OpenCV installation instructions if a newer version is needed.

  • (Optional) Hector-Trajectory-Server:

    • Install to visualize real-time camera/imu trajectories: 
      sudo apt install ros-[DISTRO]-hector-trajectory-server

2. Installation

  • Build the ORB-SLAM3-ROS package: 
    cd ~/catkin_ws
catkin_make

Part 3: Automating Node Launch with a Bash Script

The provided Bash script automates the process of launching various ROS nodes and simulations for the Spot robot setup. This script simplifies the process of starting the simulation and necessary nodes, ensuring a smooth and integrated experience.

Using the Bash Script

  1. Location of the Script: Ensure the script play.sh (or the name you've given it) is located in a suitable directory and has execute permissions. If not, use chmod +x play.sh to make it executable.

  2. Running the Script: Execute the script by navigating to its directory and running:

    ./play.sh

This will sequentially start the necessary ROS nodes and open new terminal windows for each component.

Recording and Visualizing Odometry Data

Recording Odometry Data

Use the following command to record odometry data from the /odom topic into a ROS bag:

rosbag record -O my_odom_data.bag /odom

Visualizing Odometry Data The odometry_visualizer.py script visualizes the robot's odometry data. It plots the trajectory based on the X and Y positions recorded in the ROS bag file.

Run the script after recording the data:

python3 odometry_visualizer.py

Using Real Robot Odometry Data

If you have odometry data from a real Boston Dynamics Spot robot, you can visualize this data using a similar approach.

Running the ROS Bag with Real Robot Data

To visualize the real robot's odometry data, use a pre-recorded ROS bag file named real_spot_data.bag:

rosbag play real_spot_data.bag

This command plays back the recorded data, which can be visualized or processed further as needed.

Testing on the Actual Robot

To run tests on the actual Boston Dynamics Spot robot, use the following launch command:

roslaunch spot_config simulated_spot.launch

This launch file is configured to interface with the real Spot robot, allowing you to see the tests or experiments I did in a real-world environment.

Credits and References