Turtlebot 3: SLAM and Navigation

Written by DK Kang

UPDATE:

This is a still on going project. There are some future work to be added

A very simpe navigation algorithm is implemented, and it experiences setting the goal outside the reachable distance.
This will be tested on a real robot outside the house enviroment gazebo
Ultimately, the goal is to make a robot to navigate any sort of enviroments including ramp without collision

OBJECTIVE

The goal is to use the Turtlebot 3 to autonomously map an environment and then navigate within the map using the slam_toolbox

REQUIREMENTS & SETUP

REUQIRMENTS

This project requires ROS
This project requires either a real robot of Turtlebot 3 or a gazebo simulation enviorment and its info can be found in the following link: https://emanual.robotis.com/docs/en/platform/turtlebot3/overview/
If you are using a real robot, refer to to following website: https://emanual.robotis.com/docs/en/platform/turtlebot3/setup/
we will be using turtlebot 3 packages that ROBOTIS has already developed. Refer to the following website: https://github.com/ROBOTIS-GIT/turtlebot3

setup

I will be using ROS-noetic version.

sudo apt-get update
sudo apt-get upgrade
install ROS-noetic from following website: http://wiki.ros.org/noetic/Installation/Ubuntu

install pythpn3rosdep

sudo apt install python3-rosdep
sudo rosdep init
rosdep update

install some dependencies sudo apt install python3-vcstool git gitk python3-pip catkin-lint
install slam-toolboxby runing sudo apt install ros-eloquent-slam-toolbox
install ROS_navigation by running sudo apt install ros-noetic-navigation
go to yout home dir `cd ~'
Create workspace: mkdir ws and cd _ws
mkdir src and cd src
git clone https://github.com/ME495-EmbeddedSystems/homework04-rubberdk.git
cd .. to cd into your ws dir and catkin_make
edit your ./bashrc and add export TURTLEBOT3_MODEL=burger in the file. If we are not using burger, change the model name accordingly

Nodes

This project include one node auto_nav used for autonomous mapping & navigation

This node generates xy-coordinates that move_base uses for the navigation of the turtlebot. goal coordinate points are generated by a simple algorithm that adds random numbers in between -0.5 and 0.5 to the previous goal coordinates allowing the robot to move slowly and spontaneously.

SUBSCRIBER: + map_sub (/map) ~ information about the robot's current knowledge of the environment type = nav_msgs/OccupancyGrid

MAPS

The maps directory contains two sets of map data.

map.yaml and maps.pgm were mapped by mannual control of the turtlebot
expl.yaml and expl.pgm were mapped autonomously by the turtlebot

REUSLTS

make sure to source your workspace by source ~/ws/devel/setup.bash

Demo gif files are fast-forwarded for your convinence. Refer to the YouTube link for the full video.

Autonomous SLAM & Navigation

roslaunch hw4 auto_nav.launch

This launch file makes the turtlebot to autonomously navigate and do mapping

Full Video: https://youtu.be/ceaA-sR2w_s

dokkev/Simpe-Autonomous-Navigation-with-SLAM