Hello Guys,
I have a KRC (4) of a KR60-3 kuka robot. I want to export from this program the kinematics of an execution cycle in a text file (.txt, xml and so on ..).
How do I do that?
Thank you!
Hello Guys,
I have a KRC (4) of a KR60-3 kuka robot. I want to export from this program the kinematics of an execution cycle in a text file (.txt, xml and so on ..).
How do I do that?
Thank you!
Export from what program? Are you trying to record robot motions, or have the robot import and execute motions from an external file? Your question, as stated, makes little sense and provides no details.
Sorry.
I've got a workvisual project (.wvs) with all the file structure (.src and .dat), and I've got Kuka Sim Pro.
I'd like to run this program on the simulator and to get information(angles, positions) about the joints for every simulation step.
My idea was to use OfficeLite with Kuka Sim in order to load and run the whole program structure just like with the real robot and then find a way to extract the data described above.
Is there a way to do that?
Thank you!
By "every simulation step," do you mean every programmed point, or do you mean the motions the path planner fills in between programmed points? That is, for a program that consists of P1 and P2, do you want just the data for those two points, or do you want the simulation to record data every X mm or ms along the path between P1 and P2?
Do you want to record Cartesian data or Axis data?
I want the simulation to record data every X ms along the path between P1 and P2.
I need Axis data.
Thank you!
You will need to write some KRL code for that and run this in office lite. Storing positional or joint data would require generating arrays sufficiently large. Alternatively you can write data into TXT or CSV or XML file using CWRITE.Sim pro would in this case be optional... just to see motions.
What kind of KRL code I need to write in order to run it in OfficeLite and store joint data?
Thank you!
The logging would probably have to run in the SPS, for motion-asynchronous recording. Probably use Triggers in your motion program to start and stop the logging.
Writing to a text file every few ms might run into timing problems, depending on how fast you try to push it -- you'll have to experiment with that. Might work better to log to a local array for a given motion, then at the end of the motion dump the entire array to the text file.
For the text-file writing, look up "fgets", "fputs", and "fwriteln" in the forum archives.
Thank you very much for the help,
I didn't knew the subroutines!
Do you know if "fgets", "fputs", and "fwriteln" functions work in Kuka Sim without OfficeLite? I'm trying to get any sort of output but it doesn't seem to work (also, there was a KRL printf thread here, but it seems it has been moved to a locked section or deleted).
You need office lite. The real code that comes from a real robot only can be loaded into office lite.
Thank you!
Can enabling tracing by WorkVisual help me?
Hi all, I'm working with gpatruno on this task, I'd like to specify that he means the usage of trace oscilloscope (opening then the trace with workvisual and trying to export to xml), which should be available on the kuka controller (and hopefully in office lite too). We don't have the access to the real robot, and we would like to know if it's possible to simulate this behaviour with KUKA.Sim full before buying the license.
I've never used KUKA SimPro, but as I understand it, it only provides physical simulation of the robot (and tooling, and obstructions). As such, it does not run KRL code, or provide options like the O-Scope. However, it should provide accurate motion modelling.
OfficeLite, on the other had, provides no physical simulation, but does provide actual KRL code execution, including options like the O-Scope (with the caveat that certain hardware-dependent elements, like the Fast Measurement inputs, or RSI actions, cannot be simulated by OfficeLite). While you can, technically, monitor motion in OfficeLite, those motions are apparently not very accurate.
When properly linked together, OfficeLite and SimPro are supposed to provide a nearly-complete simulation of both KRL code execution and the actual physical motion. Done properly, it should be possible to create and debug an entire set of programs in this simulation, and deploy onto the real-world robot with minimal risk of issues (as long as the simulation of the physical conditions is accurate, of course).
I've never used KUKA SimPro, but as I understand it, it only provides physical simulation of the robot (and tooling, and obstructions). As such, it does not run KRL code, or provide options like the O-Scope. However, it should provide accurate motion modelling.OfficeLite, on the other had, provides no physical simulation, but does provide actual KRL code execution, including options like the O-Scope (with the caveat that certain hardware-dependent elements, like the Fast Measurement inputs, or RSI actions, cannot be simulated by OfficeLite). While you can, technically, monitor motion in OfficeLite, those motions are apparently not very accurate.
When properly linked together, OfficeLite and SimPro are supposed to provide a nearly-complete simulation of both KRL code execution and the actual physical motion. Done properly, it should be possible to create and debug an entire set of programs in this simulation, and deploy onto the real-world robot with minimal risk of issues (as long as the simulation of the physical conditions is accurate, of course).
Thank you for the info. Just for the record, KUKA.Sim (even the trial version) CAN actually run KRL code, but you cannot load more than a single .src file (with the respective .dat) at a time. You can also import Home positions and Tool data from the config file, but after all you can't load and run the entire program structure.