[JeffNickel]

Hi All,
I am trying to import a series of points with tool orientation from a .csv file. Initially the orientations are zero degrees.
Eg: 10,10,10,0,0,0
      11,10,10,0,0,0
      ... etc.

If use "import points" and the points are displayed but I get a "can't find a start point" error.

If the list of points do not have an orientation:
   10,10,10
   11.10,10
   ... etc
Then there is no problem following the points.  If there another way to import points with tool orientations?

Also, how can I select the points for the Point Following Project without selecting one at a time?

Thanks,
Jeff

[Jeremy]

Hi Jeff, 

According to this:

Initially the orientations are zero degrees.
Eg: 10,10,10,0,0,0
      11,10,10,0,0,0
      ... etc.

Points need to be imported in this format:
x y z i j k
Where X Y and Z represent the position
and where i j and k represent the orientation of the tool normal as a unit vector. These are not Euler angle in degrees. 

If you remember, a unit vector means that square_root(i^2+j^2+k^2) = 1.
Therefore [10,10,10,0,0,0] is not a valid entry. 
For example, if you want the normal pointing in the Z+ direction, you should enter [10,10,10,0,0,1].

Hope it helps. 
Jeremy

[JeffNickel]

Hi Jeremy,
Thanks for the clarification.  This normal unit vector defines the direction the tool TCP aligns with but this does not include any rotation along the z-axis of the tool.  If, for example, I wish my to rotate (along the z axis of the tool) the tool to avoid the tool hose interfering with the fixture I cannot define this orientation in the list of points.

Once the points are imported, RoboDK converts the i,j,k values into A,B,C values. Are these ABC values then calculated to orientate the tool to the easiest accessible position?

Once imported I can rotate the tool in the simulation and export the program.  Is there a way to import points in a XYZABC format and skip this step?

Thank you,
Jeff

[Jeremy]

Hi Jeff,

You are right that this format does not include rotation around Z.
This is on purpose to give RDK some "freedom" to use its automated singularity and joint limits avoidance.
You can define the rotation around Z in your "Point Follow Project" by defining the "Path to tool offset".
You can also lock the singularity avoidance algorithm by settings "Allow a tool Z rotation of +/-" at Zero.

If this doesn't give you the result you need, let me know, there are other things we can do.

Jeremy

[sasut]

Hi Jeff,

You are right that this format does not include rotation around Z.
This is on purpose to give RDK some "freedom" to use its automated singularity and joint limits avoidance.
You can define the rotation around Z in your "Point Follow Project" by defining the "Path to tool offset".
You can also lock the singularity avoidance algorithm by settings "Allow a tool Z rotation of +/-" at Zero.

If this doesn't give you the result you need, let me know, there are other things we can do.

Jeremy

Hi Jeremy,

I am dealing with a similar problem. I import a series of points. Each point has a different plane orientation (no only Z axis, but also X and Y). Therefore, the X and Y orientations of the TCP needs to be different at each point. Can you recommend me a solution for that?

I think that the ideal solution would be importing Frames from Rhino, just like importing Points. So that they will all have unique X and Y axes, just like a typical Target created in RoboDK. But I couldn`t find out how to do this.

Thank you in advance!
Serdar.

---

Hi Jeff,

You are right that this format does not include rotation around Z.
This is on purpose to give RDK some "freedom" to use its automated singularity and joint limits avoidance.
You can define the rotation around Z in your "Point Follow Project" by defining the "Path to tool offset".
You can also lock the singularity avoidance algorithm by settings "Allow a tool Z rotation of +/-" at Zero.

If this doesn't give you the result you need, let me know, there are other things we can do.

Jeremy

Hi Jeremy,

I am dealing with a similar problem. I import a series of points. Each point has a different plane orientation (not only Z axis, but also X and Y). Therefore, the X and Y orientations of the TCP needs to be different at each point. Can you recommend me a solution for that?

I think that the ideal solution would be importing Frames from Rhino, just like importing Points. So that they will all have unique X and Y axes, just like a typical Target created in RoboDK. But I couldn`t find out how to do this.

Thank you in advance!
Serdar.

Addition: I also tried saving the points as XYZWPR coordinates and do:
Tools > Run Script > ImportCSV_XYZWPR
then I receive the attached error (ImportCSV_XYZ method works fine though).

[Jeremy]

For the first question, can you provide a bit more details, potentially some print screen of what you want to do? Because I'm not sure I get it.
If you do import point from Rhino, you should have 2 axes fixed, X and Y, only Z would be free.

You can then use that in your Curve Follow Project (CFP) or Point Follow Project (PFP). You can retrieve the points the robot follow in a CFP or PFP and convert them in target by:
Activating a reference frame -> "Right-click" the generated program from a CFP or PFP -> "Show instructions" -> Select the first Move -> Press and hold "Shift" -> Select the last Move -> "Right-click" one of the selected targets -> "Select targets"

For the second question, you need to have a frame and a tool created and activated in your station for that algorithm to work.