Hi everyone,

Here are questions we were asked during the vision system training about the presentation on the hardware of the Robotiq Wrist Camera 

Feel free to ask other ones you may have in mind!

My question has 2 parts
With the camera and ft sensor, I know it's designed mainly for pick and place. Can it be used to look at one part and focus on a specific feature on that part?
Also can the camera and ft sensor be used with something other than the gripper, such as a drill, torque gun, screwdriver, etc.?

Are there any plans to provide a more industrial connector on the end of the camera?  Since there is no locking mechanism on the USB connector, the only current option would be to pass the cable though to the control box. 

Most of our customers prefer to have some kind of quick-disconnect connector mounted on the side panel of the UR Control Box, so I would envision us needing to cut the cable in the middle, and add a connector on either end.

@bsawlor

The Wrist Camera kit for Universal Robots comes with a tool flange that recreates the UR bolt pattern (ISO 9409-1-50-4-M6). Thus, you can easily mount the end effector you need for you application. 

As of today, the Wrist Camera is mainly designed for part detection to do, for instance, pick and place operations.

What would you like our Camera to be able to perform?

Curious about the possibility whether it can now or in the future, look at a part, be programmed or taught to detect the location of a specific feature, and move there. Whether in something like a screwing app, torque, sealant, or what not. Also just thinking about it, move to the specified feature and start a recorded path, for instance in a sealing app or something similar.
I know this is, at the moment, meant for pick and place, the potential for other apps could be interesting.

@bsawlor

You could actually do this kind of program with today's version. The Vision system will detect the part and, when you teach movements to the robot, you can make them relative to the part's frame. Therefore, when programming, you could place the robot's effector at a specific place over the part and then teach it a path - an make it a relative motion. You would just need to make sure that your object has specific features so that the Camera is able to detect it and its orientation. 

This is actually a nice test we could do!

I actually tested this and the camera locate node acts like an if. So if the part is located, it will do what's inside the node.

You can then do a movement relative to the part or an absolute one.

Here is an example of the use of the camera locate node.

http://dof.robotiq.com/discussion/219/inspection-tasks-with-the-camera-detect-presence-of-elements

The only limit I can see in what you can do after the part is located is not to touch it or move it. It's because we can't control which part will be detected first nor we can tell the robot to omit one detection after it located it. So if the part remains on the workstation, the camera can locate the same forever.

But all worth to test and re-test! Does that make sense?