Automation in the meat industry: A new level
AI- and robot-based labeling of Serrano hams
Until recently, labeling Serrano hams, which is a must because it determines their progression through different aging and drying processes, was a laborious task. Since a robot could not detect the bones in the ham, human involvement was indispensable. Now, a Spanish systems integrator has developed an AI-based robotic system capable of performing this task. The labels – up to 900 per hour – are injected by a Stäubli SCARA robot designed for hygiene-sensitive applications in the food industry.
It takes 10 to 18 months of drying and ripening for a delicious Spanish Serrano ham to be ready for consumption, and this process consists of several steps. A mediumsized Serrano producer can process
54 FDPP- www. fdpp. co. uk more than 5,000 hams per day, so labeling them with the week, year, and batch number is a small but crucial step early in the process. For human operators, this entails physical strain. It also requires expertise, as they must avoid areas with bones, which the application tool cannot penetrate.
No chance for automation?
It is obvious that given these conditions, automation would be progress. Yet, although the Serrano industry is not small, this goal has not been achieved, and it is equally obvious why: As a natural product, the position of the bones in the hams, which account for 30 to 40 % of the total weight of an 8 to 12-kg Serrano ham, differs. For each individual ham, the operator must make individual choices based on their experience.
So, is there no chance of robotbased automation here, even though robots are certainly capable of applying labels? Timpolot, an automation expert located in Olot, a center of meat production in Eastern Spain, has found a solution by combining a Stäubli robot with a fastener applicator and an AIsupported vision system.
Robot-based labeling – with individual positioning
Jordi Bassols, founder and General Manager of Timpolot, describes the process:“ The hams are manually placed on a conveyor belt in a random manner. A vision system identifies the position of each piece and, using AI, determines the ideal labeling point while avoiding bone impact.”