I added ABB uploader to OP attachment.
The scope of the uploader is very similar to Kuka uploader. You can load one .mod at a time but you can load other modules one by one and merge them into the existing program. When loading multiple modules load your main module last so that the subroutines exist for call references.
For DO/DI signal mapping you can use robotWrapper.SignalOutMapping and robotWrapper.SignalInMapping dictionaries. So add real_signal:sim_signal pairs (e.g. “doArcOn”:101) to those dictionaries to customize mapping to suit your cell. Without dictionary uploader tries to find signal index from the name (e.g. do12 => 12).
Uploader loads base and tool data from module and overrides current data. If this is unwanted just delete readBaseToolData() call on upload() function. The most common thing that can go wrong in the translation is the base/tool frame location. Uploader tries to find suitable node to attach the frame to but this requires that the kinematic structure on the simulation model matches the real device exactly. Mechanism names must match (e.g. positioner component name could be “STN1”). Also be aware of tracks as the robot world frame might not be at the same location as on the real cell when using tracks. In problem cases set frames manually and block readBaseToolData().
External axis value mapping can be changed on robotWrapper.ExAxisMapping. For example if the real cell has track as axis 7 (~E1) and positioner as axes 10 (~E4) and 11 (~E5) and on the sim model track is E1 and positioner is E2 and E3 then the ExAxMapping should be { 0:0, 3:1, 4:2, 1:3, 2:4, 5:5 } using zero-based indexing.
MoveC is loaded but only as two lin motions. Few comments surround that to notify that original motion was actually circular.
-k