Checking service is advertised under /erc_aruco_score. It is defined in the erc_aruco_checker package (available here). The request message type is of ErcAruco.
It is defined in the erc_aruco_msg package as 14 float32 arrays of 3 numbers, which correspond to 14 tag positions (tags from 1 to 14). On the scene. The returned type is a single float32 value with your obtained score. The tag placed inside the box on the panel is not taken into account.
The ground truth is defined as transformation from /base_link frame of the robot to the tag name.
If you want to test your solution against the checker, you can clone the repository which consists of the checker node, config files and launch file with parameters.
Launch file erc_aruco_checker.launch starts erc_aruco_checker.py node. It takes two parameters:
- sim - boolean value if the checker is run against the simulated environment or on the real robot,
- tolerance - double value, which specifies tolerance of user detected position with respect to the ground truth.
You should call the service from within your code. The example of how to call a custom service with Python is pictured in the node example_call.py in the erc_aruco_checker package. The difference between real and simulated environments lies in the ground truth of the aruco tags placed in the simulation versus the tags in real life.