Etienne_Samson
@olivertan @Tyler_Berryman there is no embedded feature for that, but you could indeed have a set of instructions inside the Camera Locate node that will use both force sensing and Gripper object detection to know if you had a collision and to know if you picked the object. I would do something like:
You should also have a part feeding system, so if you detect a collision or miss a pick, you could shake the parts before the next attempt.Move to snapshot position Camera Locate While Force reading < collision threshold move to approach point else go back to Camera Locate Close Gripper if object is detected go to exit & place else go back to Camera Locate
Question from @olivertan during the vision system training :
Can you use object detect to compensate for the fact that the gripper does not have enough space around an object? Or use FT sensor information to detect unwanted collision?