[Keisuke.Sakai]

Dear RoboDK Support Team,


I would like to ask for your guidance regarding a tool calibration warning we are seeing in RoboDK when using a FARO laser tracker.


Setup summary:
- Measurement system: FARO VantageE
- Target: Single SMR
- SMR mounting: on the tool Z-axis, approximately 160 mm from the robot flange
- Task: calibration / TCP definition (position-based, 3 DoF)


Observed behavior in RoboDK:
- The TCP position (SMR center) is stable and repeatable.
- However, RoboDK shows:
  - “Nominal distance = 0 mm”
  - The tool appears in a red / warning state
  - After running the adjustment, even if it becomes green and the error is 0, pressing “Update” shows a message like “Valid calibration axes are missing”.


Our current understanding:
Because the SMR is located on the tool Z-axis, rotation of Axis 6 produces almost no observable XYZ displacement of the SMR center. Therefore RoboDK cannot observe an effective lever arm from wrist rotation and evaluates the nominal distance as 0 mm, even though the TCP position is consistent.


Questions:
1) Is our interpretation of the “Nominal distance = 0 mm” warning correct for this configuration?
2) For laser tracker–based tool calibration in RoboDK, is it required that the SMR be offset from the tool axis so Axis 6 rotation produces measurable positional displacement?
3) Do you have a recommended minimum offset distance (or an example mounting geometry) for using a single SMR?
4) If the TCP definition is stable but calibration axes are considered insufficient, is it acceptable to ignore the warning (use it for TCP definition only), or should the configuration always be changed for proper calibration?
5) If there is documentation or a known example for FARO VantageE + single SMR, could you share a reference?


If needed, we can share screenshots of the warning message and photos of our SMR mounting.


Thank you very much for your support.


Best regards,
Keisuke Sakai
IDEC Factory Solutions (Japan)

[Albert]

I believe the main issue with this project is that the SMR target happens to be attached too close to axis 6 (the nominal Tool X and Tool Y values are too close to 0 mm). You should move the target a bit further away from axis 6 to properly identify the robot flange. RoboDK finds the TCP based on identifying the intersection between the virtual axes 5 and 6 and applying the offset by the nominal distance.

Just applying an offset of 20 mm is enough. This can be done temporarily only for the tool measurements that move joint 6.

You can troubleshoot measurement issues by simply copy pasting a group of 3xN points.

You can find more details here:
https://robodk.com/doc/en/Robot-Calibrat...quirements

Can you share a picture of the calibration setup? More precisely, how the SMR target is attached on the robot end effector.
   

[Keisuke.Sakai]

Hello,

Thank you for your explanation. It was very helpful.

Based on your advice, we have modified the SMR setup as follows:
- The SMR target was originally mounted almost on the axis 6 centerline, approximately 160 mm along the tool Z axis from the flange.
- We have now applied an X/Y offset of about 20 mm from axis 6, only for the tool measurements that involve joint 6 rotation.


After applying this offset:
- The tool calibration result is now shown in green.
- The estimated Tool XYZ values are close to our expected values.


However, we still observe the following:
- The base calibration is shown in red.
- After running the adjustment, the result remains “insufficient pose coverage” and the final error is 0.
- The result does not change compared to before the offset.


We have attached a picture of the calibration setup, showing how the SMR target is mounted on the robot end effector.


We would also like to ask for some clarification:
1. Is it expected that the base remains red even if the tool calibration is correctly identified?
2. Could the “insufficient pose coverage” be related to how the measurement poses were taken?
3. Regarding your comment about “copy pasting a group of 3xN points”:
   - Could you please explain what this means in practice?
   - For example, does this mean duplicating an existing set of measurement points within the same measurement group for troubleshooting?


Thank you very much for your support.
We look forward to your advice.


Best regards,
Keisuke Sakai

[Albert]

The new project looks better and you took proper measurements this time.

However I see 2 separate topics that need to be clarified/fixed:

1. The base setup looks good even if it is red: When the measurements data is shown in red is because a tolerance is exceeded somehow. In this case, there is a default 2 mm tolerance that is exceeded when measuring the nominal distance between axes 1 and 2. You can change this tolerance here: Select Tools-Options-Accuracy and set Distance toelrance for warning (mm) to 4 mm. Changing this tolerance has no impact on the results or how RoboDK uses this data. This is simply to show a warning.
   
2. You used the Set Measurements reference option whe using the laser tracker only for the calibration measurements (you didn't use it for the base setup and tool setup). I strongly recommend you to not use this option unless moving the tracker during the calibration procedure is a requirement. I manually edited the calibration data you took to remove the base frame definition so it uses the tracker reference (the first line of the CSV meausrement data should be all zeros).
   

I attached your project with the correct calibration data. This calibration with these 2 edits should be valid. In your case for this project we can see a position improvement from 3.7 mm to 0.7 mm. 

See attached images.

On the other hand, for a robot like the KUKA KR 50 R2500 (50 kg payload, 2500 mm reach and 0.050 mm) I believe we could get even better results (0.4 mm position error probably). Here are some ideas that can help you troubleshoot if you want to repeat the calibration to see if you can obtain better results:

1. Did you warm up the robot and the laser tracker?
   
2. Did you wait enough time for the stabilization before taking the measurement? I would recommend 5-8 seconds. You may be able to feel the robot vibrate after it reaches a target.
   
3. Is the robot in a warehouse environment? If so, is the laser tracker close to air currents or temperature variations? Laser trackers are very sensitive to air currents and temperature changes. The measurement can vary by 0.1 mm even if the repeatiblity of a tracker should be under 0.02 mm.