Skip to content

working_environment.md

Latest commit

 

History

History
219 lines (152 loc) · 9.68 KB

working_environment.md

File metadata and controls

219 lines (152 loc) · 9.68 KB

Table of Contents

Working Environment

This workspace is being developed since 2020 and it was tested on several computers. The last tests occurred on Jul 9, 2024, on an old laptop with a 4th Gen Intel i7 Processor, and it worked on the first attempt. It assumes that you have previous experience installing operating systems on your personal computer and understand tools such as git, curl, bash, and tmux, specifically how to run commands in a terminal.

Disclaimer About Docker and ROS

All simulations in this workspace are intended to run on native Ubuntu 20.04. While many students and professionals might prefer to use Docker and install ROS in a container, I believe that the additional system architecture layer introduced by Docker is detrimental to understanding, mainly because ROS itself can be very complicated for the inexperienced explorer, as I’ve explained here. Another reason is that when working with mobile robots in simulations, we often need to visualize what is happening. Unfortunately, there is no easy way to make the graphics card driver communicate with the native system if they are different (e.g., a container running Ubuntu 20.04 with a host machine running Ubuntu 24.04 with different glibc libraries), which results in very clunky simulations and extremely low fps.

Steps to Install Ubuntu 20.04

In summary, you would need to follow these steps:

  1. Download Ubuntu 20.04 image from here
  2. Use a USB Startup Disk Creator, such as Rufus to create a bootable device using the image downloaded from step 1.
  3. Restart your computer, boot from the pendrive, and follow the instructions.
  4. After finishing the installation, boot in Ubuntu 20.04 and login with your user name and password.
  5. Ensure that you have a graphics card good enough to run this workspace packages and that its driver is properly installed.

Steps to Install ROS 1 Noetic

In summary, you need to follow these steps:

  1. After loggin in Ubuntu, open a terminal.

  2. At the terminal install the following package and add an entry pointing to ROS Jazzy repsitories at your sources list with the following commands.

    sudo apt install curl
sudo sh -c 'echo "deb http://packages.ros.org/ros/ubuntu $(lsb_release -sc) main" > /etc/apt/sources.list.d/ros-latest.list'
curl -s https://raw.githubusercontent.com/ros/rosdistro/master/ros.asc | sudo apt-key add -
sudo apt update

  3. Install ROS 1 Noetic with the following command. In particullar, I preffer installing the full package for research purposes to reduce the chances of unmet dependencies when testing and working with robots.

    sudo apt install ros-noetic-desktop-full

  4. Add an entry at the start of your .bashrc file to make your system be able to see where ROS programs are installed and the resources provided in this workspace.

    sed -i '1i export GAZEBO_MODEL_PATH = PATH_TO_MODELS' ~/.bashrc
sed -i '1i export GAZEBO_RESOURCE_PATH = PATH_TO_RESOURCES' ~/.bashrc
sed -i '1i source /opt/ros/noetic/setup.bash' ~/.bashrc

    the starting of your .bashrc file should look something like this

    source /opt/ros/noetic/setup.bash
export GAZEBO_RESOURCE_PATH = PATH_TO_RESOURCES
export GAZEBO_MODEL_PATH = PATH_TO_MODELS
# ~/.bashrc: executed by bash(1) for non-login shells.
# see /usr/share/doc/bash/examples/startup-files (in the package bash-doc)
# for examples

# If not running interactively, don't do anything
case $- in
    *i*) ;;
      *) return;;
esac

    In this case, your MUST replace PATH_TO_RESORRCES with the path of the ROS-Noetic-Multi-robot-Sandbox/gazebo_resources/worlds and PATH_TO_MODELS shoudl be ROS-Noetic-Multi-robot-Sandbox/gazebo_resources/models from this workspace. This allows both, Gazebo 11 and rviz to see the available resources and load the models accordingly.

Steps to Install and Configure Tmux and Git

This project rely on Tmux. Therefore, you MUST install it to work with multiple robots and have a clean visualization. I highly recommend to install the theme described here.

  1. Install Git hub program to be able to clone and push into remote repositories.

    sudo apt install git

  2. Install tmux program with the following command.

    sudo apt install tmux

  3. Install tmux plugin manager for a better experience.

    git clone https://github.com/tmux-plugins/tpm ~/.tmux/plugins/tpm

  4. Copy the following configuration into a file named ~/.tmux.conf.

    set -g @plugin 'o0th/tmux-nova'

set -g @nova-nerdfonts true
set -g @nova-nerdfonts-left 
set -g @nova-nerdfonts-right 

set -g @nova-segment-mode "#{?client_prefix,Ω,ω}"
set -g @nova-segment-mode-colors "#50fa7b #282a36"

set -g @nova-segment-whoami "#(whoami)@#h"
set -g @nova-segment-whoami-colors "#50fa7b #282a36"

set -g @nova-pane "#I#{?pane_in_mode,  #{pane_mode},}  #W"

set -g @nova-rows 0
set -g @nova-segments-0-left "mode"
set -g @nova-segments-0-right "whoami"

# List of plugins
set -g @plugin 'tmux-plugins/tpm'
set -g @plugin 'tmux-plugins/tmux-sensible'
set -g mouse on

# Initialize TMUX plugin manager (keep this line at the very bottom of tmux.conf)
run '~/.tmux/plugins/tpm/tpm'

  5. Open tmux by typing tmux in a terminal.

  6. Enter command mode with ctr+b.

  7. Install plugins with shift+i while in command mode.

  8. Install a patched font from here and make it default in your terminal. I also recommend turning off the Use colors from system theme.

Installing Necessary Dependencies

  1. Install catkin_tools.

    sudo sh -c 'echo "deb http://packages.ros.org/ros/ubuntu `lsb_release -sc` main" \
    > /etc/apt/sources.list.d/ros-latest.list'
wget http://packages.ros.org/ros.key -O - | sudo apt-key add -
sudo apt update
sudo apt install python3-catkin-tools

  2. Install gmapping, teb_local_planner, and pointcloud-to-laserscan.

    sudo apt install ros-noetic-gmapping
sudo apt install ros-noetic-teb-local-planner
sudo apt install ros-noetic-pointcloud-to-laserscan

  3. Install costmap_converter standalone node, to do this create a folder in your workspace, clone its repository.

    mkdir -p ~/Documents/dev/CostmapConverterWorkspace/src
cd ~/Documents/dev/CostmapConverterWorkspace/src
git clone https://github.com/rst-tu-dortmund/costmap_converter.git && cd ..

  4. Compile and source the workspace, extending the ros as sourced in your .bashrc file.

    catkin build && source devel/setup.bash

Solving TF_REPEATED_DATA

Ros Noetic has a well known bug, where repeated TF messages flood the terminal if they are detected. For some reason this may happen, specifically in bigger projects with many third party nodes. Unfortunately, even being a well known problem, there is no out-of-the-box solution, and this might prevent you from visualize what is going on in your terminal. To solve this issue you can clone the geometry2 package, commenting the line of code responsible for this and building it yourself.

  1. Create a workspace for the tf packages and clone the project.

    cd && mkdir -p ~/Documents/dev/Geometry2Workspace/src
cd ~/Documents/dev/Geometry2Workspace/src
git clone https://github.com/ros/geometry2.git

  2. Open the file tf2/src/buffer_core.cpp.

  3. Comment the line 279 save and close the file. The line you must comment looks like the following.

    CONSOLE_BRIDGE_logWarn((error_string+" for frame %s (parent %s) at time %lf according to authority %s").c_str(), stripped.child_frame_id.c_str(), stripped.header.frame_id.c_str(), stripped.header.stamp.toSec(), authority.c_str());

  4. Compile and source the workspace, extending the costmap_converter previously sourced.

    cd ~/Documents/dev/Geometry2Workspace
catkin build && source devel/setup.bash

Install This Workspace

  1. Clone the repository.

    mkdir -p ~/Documents/dev && cd ~/Documents/dev
git clone https://github.com/multirobotplayground/ROS-Noetic-Multi-robot-Sandbox.git

  2. Compile and source the workspace, extending the geometry2 previously sourced.

    cd ROS-Noetic-Multi-robot-Sandbox
catkin build && source devel/setup.bash

Next Steps

Now your environment should be ready to run the simulations from this repository and also to help you starting your journey. Click here for the next steps.