Questions from training

eric

Will Robotiq offer a product (or make a recommendation) to do the cable management?
Could we get the TCP and center of mass information for the different combination of robotiq products? Could it be adjusted automatically when you install a UR Cap?
Could I use the camera to see what the robot is seeing remotely? My customers run their robot in the night, would be useful for them.
How does the vision system decide on which part to pick next?
Could you add the waypoint node automatically when you add a Camera Locate?

Grady_Turner

@eric   Check out this link for cable management ideas:  http://dof.robotiq.com/discussion/173/ur-cable-management#latest

TCP being adjusted by a URCap would inevitably lead to folks forgetting when they switch tools out and want to manually do it (or expect the software to just "know)..  That kind of thing already happens with several customers.

Also I am guessing the camera picks the best % match model and goes with that one first to pick

Annick_Mottard

@eric @Grady_Turner

Within a Cam Locate node, if there are many parts of the same model, the Vision system will pick the one with the highest detection success rate first. 

When you run a program with multiple Cam Locate nodes, the Vision system takes snapshots and tries to detect the parts from the Cam Locate nodes, one node after the other, in the sequence the program was built. Therefore, it starts by taking a snapshot for the first Cam Locate node. If the detection is successful, it will go and pick the object the way you programmed it. If the detection success rate is not sufficient, the program simply continues to the second Cam Locate node, takes a snapshot and if the detection is successful, it picks the part. If not, it goes on to the next Cam Locate node. This continues until the last Cam Locate node has been reached, or until a part has been successfully detected.

You will notice that if you simply program multiple Cam Locate nodes, one after the other and without any logic behind it, the program may, for instance, detect the part in the first Cam Locate, do its routine, and then look for the part in the second Cam Locate right after. If you want your program to go back to the beginning of the RobotProgram sequence after detecting a part, you can use, for instance, flags to control the sequence of your program. I think a couple of teams used this method in challenge 1 from our Vision training. 

Please comment below if you would like to know more on that topic!

JeanPhilippe_Jobin

@eric

 Could we get the TCP and center of mass information for the different combination of robotiq products? Could it be adjusted automatically when you install a UR Cap?

We could provide this info but as Grady mentioned I would not recommend it since I think this would cause confusion.

Could I use the camera to see what the robot is seeing remotely? My customers run their robot in the night, would be useful for them.

No, not right now. But this would be a nice feature. Thanks.

Could you add the waypoint node automatically when you add a Camera Locate?

I would love to but this is not possible with the tools UR gives us to develop URCaps. We have asked them few times and hope to get this functionality in an upcoming revision.

eric

Thanks for the responses.  Regarding the question "Could you add the waypoint node automatically when you add a Camera Locate?", what if you put some instructions onto the camera locate "Command" tab so that it reminds the user to do this?

JeanPhilippe_Jobin

@eric
In the official release (this Friday), there will be a program example (in the "Command" tab) that will reminds the user to insert a move node before the Camera Locate as well as how to structure the move nodes after the Camera Locate.

Unfortunately, as said previously, it is not yet possible to add buttons in the Command tab that would add waypoints directly.