TurtleBot3 Navigation with Lidar Data

Setup

Clone the lidar_navigation repo into ~/catkin_ws/src with:

# On Ubuntu PC
cd ~/catkin_ws/src
git clone https://github.com/athenian-robotics/lidar_navigation.git

Install the required python packages with:

# On Ubuntu PC
cd ~/catkin_ws/src/lidar_navigation
pip install -r requirements.txt

Compile the ROS support for Contour messages with:

# On Ubuntu PC
cd ~/catkin_ws
catkin_make

Start a Turtlebot3

Start a physical Turtlebot3 with:

# On TurtleBot3
roslaunch turtlebot3_bringup turtlebot3_robot.launch 

Start a simulated TurtleBot3 in Gazebo with:

# On Ubuntu PC
roslaunch turtlebot3_gazebo turtlebot3_empty_world.launch 

Add a maze model to the empty TurtleBot3 world with:

1. Click on the Insert tab
2. Click on one of the models under ~/catkin_ws/src/lidar_navigation/models
3. Align the maze model with the TurtleBot3 model
4. Orient the screen with Ctrl+Shift mouse movements

Launch Nodes

Launch the lidar navigation nodes with:

# On Ubuntu PC
# To publish the interpretation of the lidar data 
rosrun lidar_navigation geometry_node.py

# To drive the robot
rosrun lidar_navigation teleop_node.py

# To view the contour and centroid data
rosrun lidar_navigation contour_node.py --plot_all

# To view the walls data
rosrun lidar_navigation walls_node.py --plot_all

Stop a crashed physical Turtlebot3 with:

# On Ubuntu PC
roslaunch lidar_navigation stop_node.py

Warning: Stop the teleop_node.py node before running the stop_node.py node.

Reset a crashed TurtleBot3 in Gazebo to its original starting position with Ctrl+R and restart the geometry_node.

Manual robot control

Launch keyboard teleop control with:

# On Ubuntu PC
roslaunch turtlebot3_teleop turtlebot3_teleop_key.launch

Warning: Do not run turtlebot3_teleop_key.launch at the same time as teleop_node.py.

Node CLI Options

geometry_node.py

Option	Description	Default
--slice_size	Slice size degrees	5
--slice_offset	Slice offset	0
--max_dist_mult	Maximum distance multiplier	1.1
--publish_rate	Publish rate	30
--scan_topic	LaserScan values topic name	/scan
--contour_topic	Contour value topic name	/contour
--centroid_topic	Centroid Point value topic name	/centroid
--publish_pc	Publish point cloud values	false
--pc_topic	PointCloud2 values topic name	/pc2
--verbose	Enable debugging info	false
-h, --help	Summary of options

teleop_node.py

Option	Description	Default
--max_linear	Maximum linear speed	.35
--max_angular	Maximum angular speed	2.75
--stop_angle	Full stop angle	70
--publish_rate	Publish rate	30
--centroid_topic	Centroid Point value topic name	/centroid
--verbose	Enable debugging info	false
-h, --help	Summary of options

contour_node.py

Option	Description	Default
--pause	Pause secs per scan	0
--plot_all	Plot all items	false
--plot_points	Plot points	false
--plot_contour	Plot contour	false
--plot_centroid	Plot centroid	false
--plot_slices	Plot slices	false
--max_plot_mult	Maximum plot multiplier	1.05
--contour_topic	Contour value topic name	/contour
--template_file	Template file name	/html/plot-image.html
-p, --port	HTTP port	8080
--http_delay	HTTP delay secs	0
--http_verbose	Enable verbose HTTP log	false
--verbose	Enable debugging info	false
-h, --help	Summary of options

walls_node.py

Option	Description	Default
--iterations	Iterations per RANSEC point set	20
--min_points	Minimum number of points for a wall	20
--distance	Threshold point distance	0.025
--pause	Pause secs per scan	0
--plot_all	Plot all items	false
--plot_points	Plot points	false
--plot_centroid	Plot centroid	false
--max_plot_mult	Maximum plot multiplier	1.05
--contour_topic	Contour value topic name	/contour
--template_file	Template file name	/html/plot-image.html
-p, --port	HTTP port	8080
--http_delay	HTTP delay secs	0
--http_verbose	Enable verbose HTTP log	false
--verbose	Enable debugging info	false
-h, --help	Summary of options

RViz

Launch RViz viewer to view /scan and /pc2 values with:

# On Ubuntu PC
roslaunch turtlebot3_bringup turtlebot3_model.launch

Creating a maze with gzmaze

Instructions on creating a maze with gzmaze are here