Robot Vision System Cuts Cycle Times in Half
Since 1990, Kautenburger GmbH (Germany) has been setting standards in the development of innovative products and systems in the field of sophisticated industrial automation. Among their customers is Refractarios Kelsen S.A. (Spain), a producer of refractory components and solutions that include fire clay bricks, fireproof bricks used for lining furnaces. In the production process, these bricks are stacked onto wagons, which transport them into an oven for hardening. Such bricks can come in well over 100 different shapes and sizes. In addition, in order to optimize the hardening process, they must be stacked in very specific patterns. Afterwards, they are removed from the wagons and placed on a conveyor for further handling - a strenous task for workers that called for an automation solution.
This seemingly easy automation process proved tricky. “Stacking the bricks onto the wagons robotically is not a great challenge in terms of automation,” says Christian Kautenburger, Managing Director of Kautenburger GmbH. “However, during the processing and transport the bricks will – even if only slightly – change their shape, and their positions may shift as well. Hence, simply reversing the palletizing procedure in order to de-palletize them, has led to issues: The lack of millimetric precision means that the robot may be unable to grasp the stones as it cannot achieve a vacuum, or it can even cause the delicate stacks to tumble over. This leads to downtimes, waste and requires human intervention.”
Using Roboception’s rc_visard 160m, the rc_randomdot pattern projector and the rc_reason BoxPick Module, Kautenburger has implemented a solution that eradicated these problems: Now equipped with ‘eyes and brains’, the robotic system can detect the bricks simply by identifying their rectangular shape. It provides the robot with precise grasp points – fully independent of the brick’s type, position, angle, or orientation. With the cycle time for a pick-and-place reduced from 18 to nine seconds and literally no downtime, Refractarios Kelsen now benefits from a significantly quicker handling process. Another aspect in favor of Roboception’s 3D stereo sensor: The comparatively large baseline of 160 mm allows the coverage of the entire width of the wagon (1.600 mm) in one take. Its mounting above the AoI rather than on the robot arm enables the detection of the next grasp point to take place while the robot is still executing the previous ‘place’ process.
Christian Kautenburger says: “Our customer is very impressed: They had not expected that this seemingly simple and highly cost-efficient modification would eliminate these errors entirely, nor that the implementation would be as flawless as it was. Also, the customer appreciates the very intuitive interfaces as much as we do – it is easy to make changes and updates. Robot vision expertise is not required, neither for implementation nor updates and maintenance.“