Robots have it tough, performing tasks ad nauseam to please us humans. Most of them deal with subpar working conditions, with no bathroom breaks and total blindness. When most of us humans either naturally or by some corrective means have perfect vision, why should robots suffer and work continuously in the dark?
Let the hypocrisy end by purchasing your robot some much-needed "glasses." In about an hour, your robot can enjoy 20/20 sight with TrueView from ABB Robotics, a division of ABB Inc. This Vision Guided Robotics (VGR) system not only will enable your robots to finally see what they're doing, but also understand the visual information (see the figure).
This comprehension includes the ability to parse information about their environment, such as variations in part position, type, style, and quality, all in real time. And if you thought training your dog was easy, wait until you train a TrueView system. Poor Fido isn't going to look so hot next to a scalable robotic system that obeys commands through a simple interface.
With TrueView, you get the patented SC3DT single-camera system, which is capable of 3D guidance.
You also get eVisionFactory, a software platform that allows for the standardization of deployment, training, and support throughout the manufacturing enterprise.
The video camera captures an image and analyzes it to compute a 3D position and the orientation of the part, including roll, pitch, and yaw angles. The image is then transmitted to a control system, which in turn automatically adjusts the robot's path. The communicated information includes grasp points and approach to match the location of the given part.
TrueView is mainly intended for industrial work like automotive assembly, but robots that use it probably could assemble Legos like a champ. Flexibility isn't a problem, as TrueView can deal with variations in part position, type, and style to keep your line running without the need to retool. It also improves quality by eliminating contamination caused by manual handling and by detecting defects through inline automated visual inspection.
Reprints
