Robots have been a force in the increase in flexible automation for many years, but until recently robot vision was limited. Previously, a vision system would have required a PC with a custom frame grabber to “grab” images from cameras in addition to custom and oftentimes highly complex software to analyze the images. These systems usually required vision experts, which either had to be brought in or added to staffs to maintain the systems and to create the sophisticated algorithms necessary for image analysis. This was, and continues to be, very expensive when done this way.