State-of-the-art Vision Systems For Production Automation
The vision systems for production automation are the result of ever advancing technology
. These state-of-the-art systems require increasing amounts of bandwidth for higher resolution images and increasing performance. The cameras being used for vision systems for production automation must be small because of the environment in which they have to work. Large bulky cameras cannot be used to "see" a process when the processing cycle is getting up into the 300 cycles per second range. The smaller, the better. As the camera sizes shrink, there is no room for special functions in the camera. These capabilities must be left to the frame grabber for enhanced processing of the images that are captured. A frame grabber is part of a personal computer that has the job of converting the analog signal of an image into pixels, so that it can be stored. The frame grabber is being pushed to operate at faster rates just as well as the manufacturing rates are going up. As more and more frames per second are required in the process, the limiting factor becomes the PCI bus over which the images must move. The designs of vision systems for production automation require that there be an increase in the frame buffer memory and dual ported architectures so that increasing scan rates can be supported.As all the manufacturing rates go up there is a need for the vision requirements that are monitoring them to have increased capability as well. With the bandwidth of transmission going past 100 mbps, the connection between the parts of all vision systems for production automation have to move away from slower parallel buses to the faster serial buses. Now the standard for connecting the frame grabber and the camera is a serial type. It is cabling that has less wires, but does handle the needs of high bandwidth to move the images. There is no need for the bulky cabling of the 32-bit or 64-bit variety. Having a serial connection makes the interfacing cheaper and much simpler and shrinks the size of components to make it easier to implement in tight spaces.
Digital cameras are becoming the standard for vision systems for production automation to maintain the quicker speeds and high resolution. Older analog cameras are cheaper, and they are reliable, but the demands of speed in processing images will probably make these components a thing of the past. The digital camera has become the place for the image to be converted from analog to digital rather than passing the image to the frame grabber for conversion. This will allow the frame grabber to perform other tasks to keep the speed of processing up. As the bandwidth climbs, the analog signal becomes very sensitive to noise. This is another reason that the conversion is best be performed in the camera, so that the effectiveness of the system will remain high.
High resolution color is in demand by manufacturing automation systems, over the previously used monochrome units. The inspection of the color of products has become important so the vision system for production automation must be able to perform that function. Technology is now available to move the processing of color pixels away from the work load that has been performed by the computer supporting the vision operation. The device that performs the decoding of color pixels keeps the display of color images in real time at a very high rate of display.
Greater and greater demands are being made of the vision systems for production automation, while the component size that make up these units have to be smaller and smaller. So far the technology that gives the capabilities to these required systems has been able to support the production speeds and the quality of output that is required from manufacturing processes today. With the demands of processing speeds and quality simultaneously very high, integrators must be able to balance the sometimes opposite needs of system design with the needs of manufacturing operations. There needs to be a very defined plan for merging these two possibly opposing requirements to get an optimum vision system in place. Having a well defined plan can help to control costs in design and implementation. Prior research of the characteristics of each component will help the system designers to fulfill the needs of both areas when building the system. Lower cost and optimization is better achieved when all the attributes of the units of the system can be laid out for scrutiny before the actual building begins. Having a big piece of paper, laid out on a large table, with all the requirements listed on one side and the capabilities of each component on the other side, is a simple approach to getting this job done quickly. With so many requirements and so many abilities, it is best to have a picture of the system before it is built.
by: John Mitchell