Testing inter-robot communication; two robots "sharing" usage of a plastic cylinder. Since the robots are not secured in place, positional repeatability is not perfect; the learned positions drift with use. Sorry for the poor video quality.
Maybe -- but a much larger source of error right now is simply the robots' momentum moving them around, since only friction and gravity are holding them in place.
Maybe a program would be handy here to measure some of the error drift due to the tooth gears. For example, with a top view light carefully make a mark on a circle as a starting point, by the gripper. Now have the robot step 360 degrees, and again, and see how far it is off the mark. Have it do this again by making much smaller degree steps, and measuring how the drift adds up again that way, on average.
Maybe -- but a much larger source of error right now is simply the robots' momentum moving them around, since only friction and gravity are holding them in place.
FlyByPC 3 years ago
Maybe a program would be handy here to measure some of the error drift due to the tooth gears. For example, with a top view light carefully make a mark on a circle as a starting point, by the gripper. Now have the robot step 360 degrees, and again, and see how far it is off the mark. Have it do this again by making much smaller degree steps, and measuring how the drift adds up again that way, on average.
lvecsey 3 years ago