Before running either of the tests, make sure to set the ROBOT environment variable for the simulation:
export ROBOT=sim
Two examples of the running of Lightning are provided in the package.
Box example: The scene is a tabletop with a large box on it. The box opening is facing towards the robot, and the goal position is chosen to be inside the box.
roslaunch lightning box_test.launch
This example is somewhat easy for the Retrieve-Repair (RR) module, so you will see that this module returns the path faster than the Planning-from-Scratch (PFS) module after only a few paths are added to the library.
Table example: The scene is a tabletop with four boxes on it. The positions of the boxes are randomly chosen and a goal position is chosen around the top of the table.
To see this example, run the following command:
roslaunch lightning table_test.launch
This example is more difficult for the Retrieve-Repair (RR) module, so you will see that many paths have to be added to the library before it consistently outperforms PFS. Since the library is not deleted before/after running the examples, you can run this example multiple times and see a gradual increase in performance over PFS as the library grows.
If you desire to see more than one iteration of the test at a time, change the num_test_iterations argument in the corresponding launch file.
By default, the test is run with the right arm. To run the test with the left arm, change the arm_to_use argument in the corresponding launch file to "left".
Green points: final path
Red points: invalid section(s) of the retrieved path
White points: valid section(s) of the retrieved path
Blue points: projected points from the retrieved path to the desired start and goal
Light blue points: repaired section(s) of the retrieved path
Mangeta lines: start tree for PFS planning
Red lines: goal tree for PFS planning
Yellow lines: start tree(s) for RR planning (each section to repair has its own corresponding pair of trees)
Green lines: goal tree(s) for RR planning (each section to repair has its own corresponding pair of trees)
