- Make sure the map is correct before restarting localization.
cat ~/.pal/config/99_map_config.yaml - Restart localization to load the map. You need to do it everytime after restart.
Once the localization is running, use rviz2 to set the initialpose for the localization (2D Pose Estimate). Use joystick to move and turn the robot until the laser scan aslign with the walls on map.
pal module restart localization
- Eveything is already ready and configured. You just need to launch the tmux in Scripts/navigation and press enter.
KEEP THIS RUNNING.
cd ~/Scripts/navigation ./launch_navigation_tmux.sh ## --> check the name correctly
The navigation server scripts handles the interfece with the Orchestrator and navigation + docking. It also publishes the '/robot_pose' topic.
Depending on the interface connected (wifi or ethernet)
cd ~/EurobinBrussels/navigation/cohan_nav2
./run-docker --tag <interface>_tiago
python scripts/nav_server.py
For launching docking, launch another docker shell and run the following:
./run-docker --tag <interface>_tiago
python scripts/visual_servo.py
For changing the goals and the associated april tags for docking edit the nav_server.py script and update the following:
## CHANGE THESE
named_nav_goals = {
'table': (
0.9525,
0.294,
0.0,
0.0,
0.0,
1.0,
0.0
),
'fridge': (
2.2,
0.6,
0.0,
0.0,
0.0,
0.66,
0.74
),
}
## CHANGE THESE
docking_tags ={
'fridge' : '1',
'table' : '2'
}
Nav_goals can be obtained by reading the '/robot_pose' and the tag is the ID of the tag that is being used for each location.
For adjusting the docking distance you can update the parameter 'target_x' in visual_servo_control_node or update the following line in visual_servo.py
self.declare_parameter('target_x', 1.2) ## Offset distance in metres from the front of the tag.
For a detailed setup of all things follow full_setup.md.