April 2019

n MACHINE VISION the conveyor, guiding the robot to be there at the right time. The bag is picked up and placed at the desired position and orientation in the box to achieve the required packing pattern. The packer can handle 150 bags every minute on two synchronised conveyor paths. It can pack them into three different boxes, in five different patterns. The 3D measurements can recognise and sort different-sized bags (for example, containing mozzarellas weighing 125, 250 or 400g), mixed randomly on the conveyor. “The biggest advantage of this camera is, no matter which product you put in front of it, you get a result right away,” explains A+F’s head of r&d, Joachim Bergmeyer. “You don't have to teach it to see the object. “If you want to find a bag with a classic 2D camera, you have to think about which feature to concentrate on, and what should be detected: How do I define it? Is it always the same?”, he adds. “You might need five different programs for the same product, just because the package decoration is different.” A+F software engineer Sascha Barkei reports that the cheese-packing system is proving extremely reliable. “Because the camera reports the height, we can adapt the gripping position if the mozzarella ball is not in the centre of the bag. In addition, we are also able to control the height at which the gripper approaches the bag. In this way, we can avoid gripper and object collisions, even when object height is varying by a factor of two or more.” The Sick IVC-3D smart camera being used in the mozzarella application combines image recording, illumination and analysis. It uses laser triangulation to capture information in three dimensions, providing information on shape and distance at the same time. The camera can also detect any surface damage using tools that make height, volume and contour measurements, regardless of contrast and colour. This makes it relatively easy to perform measurement tasks previously considered difficult. When testing continuous flows of products, there is a risk that only a part of a product will be captured in one image, with the rest being recorded in the next image. The IVC-3D uses overlapping images to ensure that every item is completely captured in at least one image and that no part is excluded from analysis. n The movements of the delta robots are coordinated to ensure that every bag of mozzarella travelling along the conveyor is picked up and placed in its container Secure remote access made easy www.appliedautomation/x-stk 02920 494551 / sales@x-stk.com new Sitemanager 35xx from � Connect multiple devices simultaneously � Intelligent cloud data collection module � Easy and ¿rewall friendly internet access � Persistent store-and-forward database � Drag and drop SCADA tunnelling setup � 3 port integrated managed switch