The robot initially moves to the centre of a tile, then proceeds to scan for hints. If it finds one, it turns toward it – whether it does or not, it then starts scanning square by square. It uses a mix of initial biases and hints to produce direction scores, then proceeds in the best direction. If it’s too close to a wall, it will back up. It treats green highly, and red the same as a wall (when it’s directly in front of it). It stops upon reaching the green tile.
This section will go over how to get the robot moving through the mazes. You will need a Catkin workspace set up before you continue.
It will need the following message types:
- geometry_msgs
- nav_msgs
- sensor_msgs
- std_msgs
In order to run the robot, you need to execute three commands. These should all be in separate tabs.
The first is the Gazebo simulation (# represents a number between 1 and 3 inclusive):
roslaunch uol_turtlebot_simulator maze#.launchThe next is the talker script.
By default, this does not show the robot's view, but you can pass the --view flag to enable that:
cd catkin_ws/src/path/to/scripts
python talker.py
# or...
python talker.py --viewFinally, you can run the listener script to get the robot to move.
You can skip the initial scanning phase by passing the --skip-scan flag, but you should only use this for debugging:
python listener.py
# or...
python listener.py --skip-scan