URScript Waypoint_p With Features Migration Not updating

Technical Questions URScript
1

Hello,

I have 576 Waypoints taught (some still to teach) with a feature plane in the Pendant.

I am looping through them by putting the waypoints into an array with URScript (e.g. Pal1=[PickPos1_p, …., PickPos12_p] )

The _p loads the [X, Y, Z, RX, RY, RZ] coordinates of the waypoint in the script.

A loop in my program indexes and selects which position to drive to.

I have two machines, so I want to migrate my .urp with the URScript of arrays to the second robot.

I have retaught the feature plane.

But the program is actually still driving to the original machines “_p” waypoints.

If I just drive to the waypoint it is correct. The script is not creating the new _p after the retaugth plane.

Can anyone help or advise on how to re-fresh the Script of arrays to use the new feature planes, since there is some variance in the 2 machines builds. (Closing and opening the script still keeps machine one’s waypoints)

Kind regards,

Jack

2

I had something similar happen when programming a cell of ours. When the Waypoints were copy/pasted from somewhere else, the Move command no longer kept the Feature data with it. You can maybe confirm if this is similar behavior by saving your program and opening the file from the “Script” node. (or just open the .script program in any text editor.) Check the Move commands. You should see your Feature being referenced. If it’s not, or it’s just using p[0,0,0,0,0,0] (that’s Base) then it has lost the reference to the Feature.

3

Hello Eric,

Thanks for responding.

I think you are right that the features are the problem.

But instead of losing the feature data after copy pasting, it has kept feature data from the previous robot/Cell.

When the program runs using the script with lists it collides with the part and protective stops. But when I drive to the waypoint in the URP program, the position is perfect.

So I believe my waypoint and retaught feature is correct, but the script/feature or the cell and orginal teach is baked into the script lists, or the variables _p are being memorized.

If there was a way to update the Waypoints teaching or remove and update the _p variables to the new feature, that would save me so much time tapping into the pose menu and scrolling through 576 waypoints and selecting their tags.

Kind regards,

Jack

4

Yeah it sounds suspiciously like the issue I had. When I called UR, they had said they hadn’t heard my problem exactly, but rather the opposite (yours) where Feature data was being retained when it shouldn’t have been.

My observation was the same as yours: The waypoint and Feature were both correct. However, if you looked at the actual script that is getting generated, the pose_trans() offset that it applies is simply NOT the offset of your Feature. And there’s no way to see WHY it’s doing that. My only course of action was (I think) to Move To the waypoint and hit “Set Position” again

1 Like
5

Hello @eric.feldmann,

That is interesting to hear how the feature is retained for you too.

I agree driving to each position will work, so that is an option.

However, I have been able to avoid the need to drive the robot to the positions.

Instead of using the waypoint_name_p, if I create the Position_List=[] in a line script.

And Select the positions from the pose list drop down menu.

From my point of view these poses are a kind of tag, and in the urp’s script translation they seem to use the new references.

It appears to use the correct pose_inv/pose_trans with the updated reference in the script.

I think(guess), because waypoint_name_p is a variable in the script it is somehow saved and kept in the programs full script.

The waypoint keeps its old coordinates and I havent been able to update it properly.

Maybe someone can see a neat way of clearing variables, but the waypoint_name_p are probably just asking for more trouble after that anyway.

Thanks Eric.

Kind regards,

Jack Nugent