We have about 20 work cells each with a Fanuc robot doing pick and place operations for a fairly large variety of parts. These are older robots and many have been mastered to the witness marks with the expected variation you get with that. Any time one of them fails and we replace it with a new robot we have to do extensive touching up of all the points for all parts in all of our programs. The accuracy we require is on the order of 2 to 3mm in any one direction to have it so that the points will still work. However, after replacement with the new robot (which will be mastered to factory counts), we are almost always outside of that threshold. Is there something we can do with our current robots (prior to any failures of course) to capture or preserve how they are currently mastered so that the new robot can be mastered to ‘match’ them and not require the point touchups. I have read about the quick master procedure, but it is usually discussed in the context of being reused on the same robot it was setup for and not on a replacement robot. Would it be possible to take of advantage of what quick mastering does for this scenario or is there some other strategy we should take?
Mastering Replacement Robots to Match Original Robots
-
pzzwhw -
February 17, 2020 at 3:05 PM -
Thread is Unresolved
-
-
Try using a User frame.
-
Some options I could think of would be iRCalibration (if replacing fairly new robots with R30iA controller or above), mastering to a fixture inside the cell (this will require you to put fixtures inside all of your cells, however this method can match an old RJ robot with a newer R30iA R30iB) or you could make some user frames on the existing cells. This way, if you replace a robot you will only need to teach the user frames. I guess.... (never had to replace a robot as a whole.)
Keep in mind, on older Fanuc robots, in order to match a program to a specific user frame you need to have Extended User frames option. On my older RJ2 robots this was an optional function. I had to remove the CPU and send it to FANUC in order for them to install the extendeded user frames. So be sure to check this first.
-
standardrobotguy Yes, I know reteaching the user frames can accomplish the same end result. The reason I was hoping to go the mastering direction though is because of the difficulty involved for us to reteach the user frames. We have 4 in each cell and one of the 4 requires the involvement of multiple skilled trades because significant tooling has to be removed to accomplish (large dies). It could be that is the ultimate answer anyway, but wanted to know if there was an alternate way to handle.
Famous_Fella Mastering to a fixture was the other option I had wondered about. I know we would have to create something to serve that purpose and make it available in every cell like you said, but probably something we would be willing to invest in if it worked.
-
If you drive the robot to a pointer/fixture, jog the axis in joint until they line up to your pointer/fixture, then type in the values into single axis master.