An amazing future development would be to have the motion capture with a certain level of ad-hoc task intelligence. For example, suppose I need to move 50 boxes onto a truck...I manually move the first box with the robot "watching", then tell it to move the rest of the boxes.