2. User Configuration¶

All parameters you need to adapt to your setup are located in the wenglorUserConfig module. Adjust them according to your needs before running the program.

wenglorUserConfig.src parameters

Parameters¶

Edit the following parameters in wenglorUserConfig.src:

Parameter	Description
`W_CONNECTION[]`	Name of the EthernetKRL channel. Must match the name of the `wenglorVision.xml` file. By default, the names already match — do not rename it.
`W_USE_CASE[]`	Defines whether the camera is on the robot or not (`camera_on_robot` or `camera_not_on_robot`).
`W_CALIBRATION_TARGET[]`	ID of the calibration plate: `zvzj001`, `zvzj002`, `zvzj003`, or `zvzj004`. Select `zvzj001` if using ZVZJ005 and `zvzj002` if using ZVZJ006 (same size).
`W_CALIBRATION_JOB[]`	Name of the uniVision job for calibration.
`W_DETECT_OBJECTS_JOB[]`	Name of the uniVision job for detection.
`W_DETECT_TARGET_JOB[]`	Name of the uniVision job for detecting the calibration plate (used by `updateReferenceFrame`).
`W_SAFETY_OFFSET_MM`	Z offset in millimeters for the validation of the calibration.
`W_NUM_CALIBRATION_POSES`	Number of calibration poses to use. At least 5 poses are required, and the value must match the number of poses taught in `moveTocalibrationPose()`.
`W_USER_COMMAND[]`	The routine to execute (`singleDetection`, `multiDetection`, or `updateReferenceFrame`).
`W_BASE_NUM`	Base number used for the reference frame in the `updateReferenceFrame` use case.
`W_BASE_NAME[]`	Name of the reference frame base (`wReferenceFrame` by default).
`W_MACHINE_POSES_TAUGHT`	Set to `TRUE` after teaching the poses relative to the reference frame in the `updateReferenceFrame` use case. See Robot Program → `updateReferenceFrame`.

Mobile platform use case¶

For the mobile platform use case (updateReferenceFrame), also set W_BASE_NUM, W_BASE_NAME[], and W_MACHINE_POSES_TAUGHT as described in the table above. For the general concept, see Wenglor Robot Server overview in the wenglor robot vision manual.

Teaching the poses¶

The movements to the robot poses are defined in the pose procedures of wenglorUserConfig.src. Add the movements to your taught poses inside them:

Procedure	Purpose
`moveTocalibrationPose(poseNum)`	Moves to each calibration pose (`CASE 1` … `CASE 5`). Add or remove `CASE` blocks to match `W_NUM_CALIBRATION_POSES`.
`moveToDetectObjectsPose()`	Moves to the detection pose used for object recognition.
`moveToSafetyPose()`	Moves to a safety pose so the operator can remove the calibration plate (`camera_not_on_robot` only).
`moveToDetectTargetPose()`	Moves to the pose from which the calibration target is detected (`updateReferenceFrame` only).

Teach a minimum of five calibration poses (more can be added for better accuracy) in moveTocalibrationPose().

Teach the calibration poses

Note

The number of taught poses must match W_NUM_CALIBRATION_POSES. To add poses, copy a CASE block in moveTocalibrationPose(), increase the pose number, and update W_NUM_CALIBRATION_POSES accordingly.

For the detection pose (moveToDetectObjectsPose()), the safety pose (moveToSafetyPose(), camera_not_on_robot only), and the target pose (moveToDetectTargetPose(), updateReferenceFrame only):

Teach detection, safety and target poses

To test the calibration poses you can call the function testCalibrationPoses(). It must be called from wenglorMain.src.

Test the calibration poses

For details on how these poses are used in the calibration and detection flow, see Robot Program.

Note

Also check that KUKA is selected in the robot manufacturer drop-down of the robot server on the Machine Vision Device website (e.g. B60, MVC). See Settings on Device Website in the wenglor robot vision manual.