Hi pros,
one quick question, I saw a video showing the repeatability of the UR robot:
https://www.youtube.com/watch?v=9orN_aUDY7w

Listed in the specifications of the UR robot is a repeatability of 0.1mm. A question that we have been asked is that the video above shows that the robot is within the specs given. What if you turn off and then on the robot and try to run the same program. Would you still be within the limits or will the robot initialisation steps make this deviate a bit more?

I would presume that it should have the same repeatability with the absolute encoders.  That would be an interesting test to run, to e-stop the robot, start it back and see if it holds the same, or to have someone interfere with the robot and see what the numbers look like after resuming operation.

@Ryan_Weaver any interest in doing that from the Axis folks?  

I'm with @matthewd92 , I think that the absolute encoders are what dictate the repeatability and there is nothing that changes mechanically after you e-stop.  The robot still knows where it is.  

I think that is proven by many customers running tasks with tight tolerance who turn off their machine every night, yet the robot hits the points it needs to when they power it up in the morning.  

Hi guys, the only time I have seen the repeatability drift is with a joint change.

I guess @Stefan_Stubgaard could explain us why repeatability doesn't drift with reboot or joint firmware update. Stefan?

@Ryan_Weaver explains it beautifully. As the system has true absolute encoders* it will remember it's position after reboot. Joint firmware doesn't affect the positioning.

* True absolute encoders = no battery backup needed

Conventional absolute encoders requires battery to maintain position information in the memory, so a Universal Robot is never affected by loss of position due to dead batteries.

@Ryan_Weaver @matthewd92 @Stefan_Stubgaard

The only time we have seen a loss of repeatability was when the robot crashed at full speed into the machine structure.  The base joint mechanically shifted in it's locking key mount, we had to re calibrate to get the program positions back 

@Grady_Turner don't you know you're not supposed to do that?

@matthewd92  It wasn't me driving when it happened! :smile: