One essential function of the vision system is to determine the position and orentation (pose) of components for flexible parts feeding. Pose estimation is performed using built-in functions of the AdeptVision software, and must be fast enough not to interfere with the assembly cycle time. A secondary function of the vision system is to register pallets and modular work tables to a robot's world coordinate system, avoiding the need for alignment hardware. Still another use may be error recovery, wherein the cameras can be used to inspect critical-points in the system, or in-process assemblies.
The vision system uses a number of standard CCD cameras, mounted either above the flexible parts feeders or the robot arms. Since the number of camera inputs to the Adept Vision system is limited to four, a low-cost, custom video multiplexer was developed, utilizing a monolithic video-switcher integrated circuit. This allows up to four cameras to be attached to each video input on the video hardware.
In keeping with the quick-changeover philosophy, the vision routines are designed to be reusable; that is a given routine may be used to locate several different but similar parts (i.e. similar asymmetries, topology, etc.). This approach has many advantages, including minimizing the number of software routines. In addition, this reusability allows for software modularity and "agility." For example, by parameterizing the characteristics that a routine searches for, it can be applied to parts that have a similar profile but are of a different size. This means that parts with similar geometries to those in the parts library can be added to the system by simply modifying the inspection procedures that call these lower-level, reusable routines. A third advantage of reusable vision routines is that they lead to object-oriented programming wherein a general recognition routine may be an object class.
Last update: 7 June 1996
E-mail: rdq@po.cwru.edu