iRVision Robot picks part with slight offset

    Hi,

    We got a system with 5 Fanuc Robots for paint work. The first robot uses iRVision to locate parts on a desk and put them on the holder for painting. Today we run into a problem after trying to learn a new part to iRVision (I did this the first time). Now all the other programms that were teached by an extern company will have a slight offset while picking the parts.

    Here is what I did:
    Made new Vision Programm
    Learned the part
    Saved the postion in VR[1] (which is also used by other programms)
    Wrote a new programm which picks the part from this position

    This task was an exercise to see how the iRVision works (adapted from here:

    External Content www.youtube.com
    Content embedded from external sources will not be displayed without your consent.
    Through the activation of external content, you agree that personal data may be transferred to third party platforms. We have provided more information on this in our privacy policy.
    , without the calibration process)

    We tried to restore the auto-backup but the mistake is still the same. I can't explain where I could change something that now this offset occurs.

    Does anybody have a solution for this problem or an approach what we can try to solve this? Tell me if you need more information.

    Thanks so much for your help!

    Best regards, dofrey

    Thanks for the fast answer. We thought about this. But somehow the offset is not constant. The error changes over time :wallbash:

    We set the reference position new and then the robot picked 10 parts properly. But after a time there was an error.
    I cant explain it, as I only added another vision process and a new program. Is the vision register that I used maybe a hint?
    Or could i change something on the calibration of the camera with making these two programms?

    Are these parts picked up by suction or electromagnets? Is it still picking the parts just not in the desired location? And are these two parts fairly close in size shape? I'm still new to the iRvision but i'm wondering if you should have taught your new position to VR2 and used that instead of the same VR as is in the other programs.

    "I could tell that my parents hated me. My bath toys were a toaster and a radio."

    The parts are picked by a mechanical gripper. The robot will "crash" into the parts as it does not hit the hole, where the gripper should go inside and gives an error. The part that I teached is the same that was already teached to exercise but in a total new program.

    Another problem is, that the camera has two positions. So there are complicated calibrations in the program that I dont understand. Thats what makes it difficult to check the VR as the positions don't tell me anything.

    dofrey

    What are the twow position of the camera ? You have to tells us everything. I feel like Steven Spielberg trying to imagine your machine
    Do yo have a problem posting the program regarding the vision. I dont care the one opening and closing the gripper

    Tell us what dont you understand, is it the logic or one particular command ?

    I dont even know how the 5 robots communicate. Are all of them separate controllers ? etc etc
    Are they suppose to do exactly the same task s ?
    Is there a Master and four slaves ? or are they all individuals ?

    Retired but still helping

    Sorry for the late answer. As you can imagine, it was a hard time when the system was not on production. But now we figured it out what was the error and fixed it!
    As the camera was on a slider moving between two positions it untied the lens of the camera from vibrations. Of course this gave a new random offset whenever the camera moved to the new position.

    So always check also hardware. Especially when the error is random.

    Thanks anyways for your help Fabian and SHIFT_Lock!

    Glad you got it figured out! Ours are fixed so we don't have to worry so much about that. Did you guys end up doing anything to try and limit the vibrations? I'm not sure how your setup is whether it's mounted to a linear bearing or just a pneumatic cylinder but hopefully you can reduce the vibration to prevent further issues down the road.

    "I could tell that my parents hated me. My bath toys were a toaster and a radio."

    As we had a lot of systems in the field that suffered from loose lenses due to vibrations, we started putting a bead of silicon on the camera across the lens assembly. This works super well and can be removed if the camera needs to be changed.

    The camera slider has rubber bumpers on the left and rigt side and one of them fell off. We actually figured it ojt because if the clicking noise that appeared. We just added this bumpers and lowered the speed of the slider by reducing air pressure.
    We will check the silicon bead idea! Thanks!

Advertising from our partners