Program repetability

what causes the robot to lose it pick position on a pallet form one time to the next?

From one time to the next… do you mean one part to the next, one setup to the next, one day to the next??

I see it quite a bit between setups. I’ll run a setup, Everything dialed in nicely, then set up something else, then go back to the previous setup and I have to adjust waypoints for the pallet and often for the vise points as well.

Yes that’s exactly what I mean. We purchased a new UR3 e series and we can get a program set and it runs perfectly and then run a different program and then go back to the first program and everything has moved!

I haven’t figured out what causes that. It’s been happening for going on 3 years on 2 different robots for me. Would be nice if I could just change templates, gripper fingers and programs and have it just work!

Might be my pallet table and/or robot pedestal are not rigid enough. Might be that it doesn’t power up the same (like when an alarm powers down the robot). When you power it back up you can see, hear and feel it move quite a bit as everything gets turned on. I assume it’s always going to the same place but who knows? Might just be the nature of the UR family. We also have a Doosan robot here but it only does one job and never needs to be changed so I don’t know if the Doosan’s have the same issue or not.

How much of a difference are you seeing? I expect anywhere from 1/32 to 1/16 inches between jobs based on what I’ve seen so far.

Subsequently, the above issues are why I bitch and complain so much about the GUI and it’s functionality and want to make it easier to do setups.

I think most people have a specific job they use their robot for so many (most?) users don’t have to change setups very much.

I use mine for many different jobs and do setups very frequently (usually at least one a week, sometimes 3 or 4).

If you are making changes to the TCP that was used to program your original points, this will definitely cause an issue. I’ve seen this more if the orientation was changed than if just the cartesianal offset was adjusted. But both will affect it. Likewise if you taught it to a Feature and have since adjusted the Feature, that will also affect it greatly.

If you’re trying to juggle between many different setups, you should consider using the Features. You can machine/laser points to your fixture as your 3 reference points, and then by taking your TCP to those same 3 spots, you should be able to reset all your waypoints by just touching up these 3 specific points.

My robot is bolted solid to the same work station that the pallet system is on, I’m sure that it’s not the physical setup. It’s changing from 1/8 to 1" in some cases. We have hundreds of different parts that use templates that fix them to the work station in a pallet
layout. the thing is that once you have programmed a certain part in that system and saved it, it should be repeatable every time!

Hi Eric,
We are talking about judging one part after robot has been programmed thousands of those parts have been run successfully and then running other programs for a while then going back to that successful program to find that it doesn’t work or doesn’t even start
in the same area.

Oof. Nope, never had that happen. You might want to call your distributor. It might need to be serviced. Could be a joint going out, or other bad calibration data. The joints are absolute encoders, and I believe the unofficial repeatability is supposed to be ± 0.5mm

I only change gripper fingers on the Hand-e grippers. The grippers themselves only come off if there’s a failure (f/t sensor, wrist, connector, splitter etc have all been reasons so far). Most of the fingers have a locating pin for the rack system, but even without the pins the fingers appear to repeat within ±.005.

I’m like 95% self-taught and haven’t done anything with the existing features that were set up. I don’t really understand how they work or how are set. We have them for the robot base, the machine zero and some for the pallet system that I don’t understand. When I try to send the robot to some of them manually it tries to go under the table LOLnotLOL.

Oh, wow. That’s massive.

Only time I saw anything close to that was right after the install of our 2nd bot… one of the 4 bolts holding the base to the stand was loose and the whole thing rocked about 1/16 inch at the base. By the time it got out to the TCP it was close to an inch off. Took us weeks to figure it out.