Hey.

You see, the problem comes at random time, but I am sure it always arrives. And it is not very easy to repeat it, as it only comes after more than 30+ cycles.

It is a simple Pick’n’Place operation, where product is picked form a conveyor (static), placed on a laser and later placed on another conveyor for further processing. Problem is that over some cycles, the robot deviates form the waypoint coordinates and eventually will not pick up product from the conveyor, and then problem escalates till it feeds up empty laser and later on second conveyor has an empty shuttle and finally AOI (automatic optical interface) fails to pass a step as the shuttle is empty (I know I know, there should be some detection upon picking up the product, but this is not the biggest problem now). What I have seen is that X an Y coordinates are the problem here, and usually after these events I have to readjust all the Waypints (total of 8). The same thing happens after emergency stop or force related stop, where robot remains in active mechanical brakes (this I could understand as these are emergency related actions, those should be stopped immediately no matter what, with this event the difference is even worse – bigger, could be up to 5mm). Still after resuming a movement is detected (as if it is looking for a home point).

I have tried to investigate this, what is odd to me:

• I have written down the coordinates of waypoints – after these events when I want to go to these waypoints (Move to Waypint -> Auto), I can see, that the numbers are the same (or a bit different, but 3rd digit after comma is nothing), but in the real life it could be several mm, even up to 5mm. As I have Pnemuatic double acting gripper, it couses robot not to pick up product or piking it up with sideways-forced movement and eventually with an error because the sideways resisitive force is too high and it has reached robot safety force limit.

Question 1 – is it possible, that over time, physical points deviates form the software saved XYZ? Couse this is what is happening here.

Question 2 – how to avoid this as this way robot can not be used independantly.

@Mikus I have no answers for you but I had the same experience a couple of time with e-series in the last weeks, usually happens after a stop or a restart and all the waypoints shifted in the same direction for me, around 1.5 mm off in Y position for everything. There was one time where it was major and I found out that my robot base was not fixed properly, all other time everything was normal and I had no other choice but to re-teach some critical points. If someone has an answer to this, I'll be really happy.

We recently had a similar issue with a UR10 CB3 following a protective stop we’d have waypoints off by as much as 3-5 mm. The issue was found to be a defective joint that was assembled incorrectly. 

I would contact your distributor/dealer and ask them to investigate the issue or get a loaner robot and determine if it resolves the issue. 

Other than that, verify that your base attachment is not moving over time. We did that by placing a dial indicator on the base to ensure it was not moving. 

Hey, thank you for your answers, but I rather think that the solution could be found by contacting the supplier. Hopefully.
Base is steady, aluminium profile (50mm) special desk with all screws in and connected to laser machine to ensure movement. This could not be the issue. Normally, if a shift was finished on good mark (no problem at the end) after switching on usually I do not see the error. It only happens during action and after E-stop, of course (regarding E-stop, there are remarks in user manual that back up this action, that there might be some deviation).

Hello all.
Just FYI, problem is solved, good thing that robot is fine, bad thing - it was my fault at the end.
Problem lies in a situation, that although mounting base was steady, there still was some degree of freedom in the base-bolt-platform relationship. It was almost impossible to see the deviation, but my local friend (dust) made a mark over time in the platform. By taking a closer look it can be seen, that there was a very small/slight movement and this was the reason of the problem. 
So, quick tip - always check base-bolt-platform fixation infrastructure as even very small deviation can lead to some problems at the tool positioning. If you think it is ok, just think twice B)
So far, working 3 days, no issue in this topic.
Thank you for your answers