Currently working on improving a couple things on our training cell (see this link to see the training cell) and I was pondering what everyone does when creating work spaces from scratch for a new application. Generally on our production equipment we don't modify the work spaces, although I have added a few basic ones sporadically. When I did create new or modify existing ones I more or less just jogged the robot around to get a approximate value for my cube. Not necessarily the most efficient way to do this. Does anyone have tips/tricks for setting up new work spaces? I have use of AutoCad and K-Roset and have been playing around with doing a rough layout in AutoCad and then creating a obstacle in K-Roset to sort of plot the Work Space in advance. Still feels like I'm doing a lot of guessing/eyeballing values. Any thoughts or suggestions would be appreciated.
Best practices when creating work spaces
-
ericwiz7923 -
July 27, 2021 at 5:37 PM -
Thread is Unresolved
-
-
I try and use the following method:
- A fabricated tool with the correct TCP to reflect the known tool, and the same in KROSET model.
- Accurate dimensioned cell layout modelling within KROSET.
- I create transparent 'boxes' to represent the actual workspace dimension using the shape generator.
- I then use the robot to 'teach the lower and upper corners (vertices) of the 'transparent box'.
- This then represents the 'actual size' and lower/upper limit values that I enter into the workspace fields.
- But instead I create a file for AUX DATA and put my workspace values in there and just load into the live robot and 'tweak' where required.
-
I'll have to go back and verify my TCP, that is a good point I hadn't thought of. Otherwise it sounds like I'm doing something pretty similar. Thanks for the feed back!