April 2019

48 n MACHINE VISION April 2019 www.drivesncontrols.com Say cheese: cameras help to pack mozzarella T he global market for mozzarella cheese was worth $10.1bn in 2017 and predicted to reach $15.1bn by 2025. More than 70,000 tonnes of the soft white cheese are produced, packaged, sold and consumed every year, and sales are continuing to grow at double- digit rates. But, mozzarella is sensitive and spoils quickly. It needs to be handled gently, even at the packaging stage. Until recently, the plastic bags used to immerse balls of mozzarella in brine were transferred by hand or as bulk materials into secondary packaging, such as cardboard boxes. With the vast quantities now being produced, this is no longer possible. Handling the bags in bulk does not make optimum use of the storage space in the boxes, and the bags’ sharp edges can damage each other. Because of this, and the fact that the bags are sometimes dropped during the handling process, there is a high risk that the bags will start to leak. Belt picking So a better process was needed to package the mozzarella bags rapidly, but gently. An attractive alternative to manual or bulk material methods is to use a belt picking system. The German specialist in end-of-line packaging machinery, A+F (Automation+Fördertechnik) came up with such a system that uses machine vision to track the bags of cheese as they pass along a conveyor and guides four-axis delta robots to pick them up and place them safely into their boxes. To ensure that the irregularly-shaped and glossy mozzarella bags are picked precisely and reliably, A+F turned to a proven machine design, called FlexoPac, and adapted it to the mozzarella application. Depending on the box feed, the packer can handle up to 40 boxes (single path) or 70 boxes (double path) per minute. The packing machine is designed to place portion cups, thermoformed trays, bags or blister packs into boxes using four- axis Delta 3 robots. It is ideal for ready- packaged products. In addition to trays and wrap-around cartons, it can also pack boxes with covers and two-part trays. Too tough for 2D When considering how the FlexoPac machine could be adapted to packing mozzarella, the designers and system integrators first looked at whether they could use 2D cameras to guide the robot picking arms. They concluded that the bags would be difficult or even impossible to detect using 2D cameras. The bags’ glossy appearance and irregular shapes – the cheese do not always lie in the middle of the bag – were found to be too complex for 2D cameras. They therefore started to test a smart 3D camera from Sick, which offered the possibility of height-based detection. The camera (the IVC-3D) detects the height of objects on the belt. It is set to find any object with a height of 4–44mm, and to extract the relevant features, which are transmitted to the robot controller using a simple, robust protocol. This data includes the object’s centre, orientation and height. Optional parameters such as area and surface inspection results can be used for sorting. Based on this data, the robot controller calculates where to pick up the bag from Packing of bags of mozzarella cheese automatically into boxes is a challenging task, requiring the delicate cheeses to be handled with care. A German packaging specialist has come up with a machine that uses a 3D vision camera to guide delta robot arms to pick up bags of cheese from a conveyor and to place them gently into boxes. The 3D machine vision camera (left) guides the delta robot arms to pick up mozzarella bags and place them into cardboard containers

RkJQdWJsaXNoZXIy MjQ0NzM=