Description
1 Overview
The objective of this lab is to get you familiar with the mapping, localization
and path planning capabilities of the ROS navigation stack. The final goal is
to write a script to make a simulated and real robot autonomously navigate to
a series of global waypoints.
We encourage you to use all available resources to complete this assignment.
This includes looking at the sample code provided with the robot, borrowing
pieces of code from online tutorials, and talking to classmates. You may discuss
solutions and problem solve with others in the class, but this remains a team
assignment and each team must submit their own solution. Multiple teams
should not jointly write the same program and each submit a copy, this will not
be considered a valid submission.
2 Lab Instructions
2.1 Gazebo Simulation
For this lab, you will be required to leverage Gazebo simulation environment can
be used for code development and initial testing. A reminder, the tutorial on
simulating the Turtlebot in Gazebo and initial modifications to the simulation
environment can be found in Lab 2. In addition to this initial setup, for this
lab, we have created a Gazebo environment similar to the one you will find in
the lab to test your navigation code. The files and directions on how to use
them are available here,
https://github.gatech.edu/swilson64/7785 Lab5 Gazebo Files
N ote: This maze environment will be identical to the one used for
your Final project. The signs on the wall are associated with Lab 6
and the Final so you can ignore them for now.
Your team will be expected to perform the lab in simulation and on the robot.
To perform this lab in simulation, follow the same steps as in subsection 2.2
and subsection 2.3, but with Gazebo running instead of bringing up the real
Turtlebot. Gazebo will subscribe and publish the same messages as the real
robot.
1
2.2 Create a map through teleoperation
Generate a map of your environment using the instructions found at:
http://emanual.robotis.com/docs/en/platform/turtlebot3/slam/
N ote: If during any of the steps your PC or Turtlebot throws
errors of the form,
ERROR: cannot launch node of type [example/example]:
example
You are missing that ¡example¿ package, install it with the
command
sudo apt-get install ros-melodic-example.
You can tab complete to find the correct name or Google the
name of the package to find the correct instillation command for
ROS Melodic.
In later stages of the lab, you will have your robot navigate to a predefined goal.
To ensure that everyone uses (approximately) the same global coordinates, start
your robot with its wheels on the blue tape marks on the floor, facing in the
direction of the arrow. Then drive your robot around to complete the map.
You can also save this file wherever you would like in step 9.4.1 by changing the
directory and file name after the -f command.
2.3 Localization, Path Planning, and Navigation
To start, use the map you’ve generated to have the robot navigate to a point
you specify in the rviz GUI. Instructions found at:
http://emanual.robotis.com/docs/en/platform/turtlebot3/navigation/
N ote: If you changes the save location of your map files, make sure
they are the same as defined after the map file:= … command
in step 10.1. This navigation is using a flavor of particle filter for
localization and A* for path planning.
After you verify everything is working in RViz, echo the topic that is responsible for handling the 2D Nav Goals you are creating through the RViz GUI
through the command,
rostopic echo /move base simple/goal
you should notice the topic is publishing the map frame it is navigating in as
well as the goal pose. The pose consists of 3 positional coordinates (x, y, z),
2
and 4 rotational coordinates (x, y, z, w). These four coordinates are a way of
representing a rotation called a Quaternion. To publish to this topic rather than
use the GUI, use the command line argument,
rostopic pub /move base simple/goal geometry msgs/PoseStamped
‘header: stamp: now, frame id: “YOUR MAP NAME”, pose:
position: x: 1.0, y: 0.0, z: 0.0, orientation: x: 0, y: 0,
z: 0, w: 1.0’
the message being published here is a PoseStamped geometry message.
https://docs.ros.org/en/diamondback/api/geometry msgs/html/msg/PoseStamped.html
You will need to use this message type to pass your scripts goal points to the
NavStack for this lab.
2.4 Drive to global waypoints
Read in global waypoints.txt and navigate the robot to each waypoint in
turn. Specifically, the file will contain 3 waypoints not known to you until
grading time. An example file is provided with this assignment, but feel free
to try other waypoints. Your code should have the robot drive to the first
waypoint, stop for 2 seconds, then go on to the second waypoint and so on. The
waypoint file lists seven values per line, corresponding to the 3D position and
the orientation quaternion for the waypoint, in that order.
3 Parameters that should be tuned
When using the navigation stack, there are a few parameters you can adjust
to get more consistent performance. All of the parameters are found in the
turtlebot3/turtlebot3 navigation/param directory, and they’re loaded by
the navigation launch file. All you have to do is change the parameters in the
various .yaml files (you can open these with a text editor), no compiling needed.
Recommended parameters to adjust:
• dwa local planner params.yaml
– xy goal tolerance, yaw goal tolerance: Increasing the goal tolerance will allow the robot to stop when it’s ”close enough” to a goal
point. This can prevent the robot spinning around in circles for a
while when it gets close to a goal to get to the exact correct point,
provided your solution can handle reaching a larger area as a goal
rather than a very specific point/angle.
3
– path distance bias, goal distance bias: Increasing these will cause
the robot to more strongly follow the planned path or move towards
the goal. Think of it like the switching controller we discussed in the
lecture.
– max vel x, min vel x, max rot vel, min rot vel: These all set
bounds on how fast the robot will move. Fast moving robots may
shake and mess up localization, so reducing maximum velocities may
improve performance.
• costmap common params burger.yaml
– inflation radius: This will determine how large to make obstacles.
Increasing this will prevent the robot from driving close to walls.
– cost scaling factor: This will determine how much the robot should
be repelled by obstacles. Increasing this will make the robot prefer
not to drive close to walls, but will allow it if necessary.
4 Grading
Show an instructor the generated map of the simulated maze
20%
Show an instructor the generated map of the maze 20%
Navigate to 3 waypoints specified at grading time in
simulation
30%
Navigate to 3 waypoints specified at grading time 30%
5 Submission
You have two required submissions for this lab.
1 Your ROS package in a single zip file called TeamName Lab5.zip uploaded
on Canvas under Assignments–Lab5.
2 A live demonstration of your code to one of the instructors. This can be
done anytime before the due date at the top of this lab. Class will meet in
the lab room on the due date to allow everyone to demo their navigation.
If you demo before the due date you do not need to come attend class that
day.
4
