I am doing an experiment where I am comparing the resulting feeding time of an officeLite Simulation and a Physical robot (KR16+KRC2) performance. I seem to get a 20% feeding time discrepancy.
In both cases I use the physical hardware for actuating grippers etc. The load (gripper+component being fed) weight is not substantial (Less than 1kg).
I have compared the ROBOCOR, MACHINE and CUSTOM files.
I am sending the movements to Robot controller (OfficeLite and KRC2) via TCP (openVarJar) and suspect the OfficeLite reacts more quickly to the coordinates sent. This leads me to think that although the movement time is correclty simulated, the OfficeLite simulation performs the "premovement" calculations sustantially faster than the physical KRC. Does anyone have any experience on this matter? The PC Running officeLite is an i7 3rd gen but the virtual PC running the software is rated at 1gb ram etc....