OMRON Industrial Automation and the meat company Prolongo-Faccsa recently presented the results of ALIMENTE21: a four-year research program that demonstrates how AI-based computer vision and predictive analytics can transform quality, safety, and efficiency in the food industry.
The project was approved and funded by CDTI in 2021 and completed in June 2025. It represents one of the most comprehensive deployments of industrial AI in the Spanish food sector to date, tackling challenges that have resisted automation for years: from verifying cleaning and disinfection processes, ensuring the absence of bacterial contamination, to detecting potential physical contaminants such as bones or cartilage in sliced products.
It is explained by Sergio Vega, Director of R&D&I and Sustainability at Prolongo-Faccsa:»Prolongo-Faccsa has more than 200 years of history in the pork sector, producing traditional products such as sausages and cured hams. ALIMENTE21 has given us capabilities that a decade ago would have been unimaginable: controlling and predicting the optimal curing point of a batch of sausage through vision, verifying proper cleaning and disinfection of surfaces and absence of bacterial contamination, monitoring for the absence of any small bone or cartilage in sliced products, among other solutions. It isn’t just incremental improvements; it’s about changing what food safety and process efficiency mean in practice.»
Applications across the production chain
- Curing Chamber: Computer-vision cameras monitor the sausages throughout the curing process, measuring how they shrink in size and weight. By combining real-time observations with predictive models, the system accurately anticipates when a batch will be ready for release, reducing curing times, energy consumption, and freeing production capacity.
- Pork Carcass Slaughter Line: Quality control on pork carcasses has traditionally depended on visual inspection by operators and veterinarians. The system developed in ALIMENTE21 automatically identifies, with very high precision, the presence of bile remnants or intestinal contents, even when contamination is minimal, and can support inspection by Official Veterinary Services.
- In the realm of equipment maintenance, a system was installed that reads variables such as current draw and vibrations of the drive motors on the conveyor chains, identifying early signs of wear far enough in advance for maintenance teams to intervene before a downtime occurs.
- Packaging: Cameras with ultraviolet illumination on the slicing lines distinguish in real time between surface dirt and dangerous bacterial biofilm. On the other hand, SWIR cameras combined with AI-based inspection algorithms detect bone fragments up to 10 mm, even when hidden between slices of Serrano ham. An additional AI vision system checks the sealing of each tray, identifying defective packaging before it leaves the plant.
The FH Series Vision System from OMRON
The solutions were built on OMRON’s FH Series Vision System, a high-speed image-processing platform that combines a wide range of cameras with AI-based inspection algorithms. The FH Series incorporates OMRON’s AI Fine Matching technology, which builds inspection models from images of correct products, without the need to manually program rules for each defect.
For bone detection in ALIMENTE21, the project used the FH Series’ SWIR cameras, which operate in the short-wave infrared spectrum and can see through the surface of food products. This is the key that enables locating fragments as small as 10 mm even when hidden between layers of sliced ham, something neither conventional cameras nor the human eye can reliably achieve at industrial production speeds.
In the words of Rubén Rojo, Marketing Director for OMRON Electronics Iberia and project lead: «Each stage of Prolongo’s production chain posed a completely different inspection challenge. The FH Series gave us a single platform — one controller, one software framework, and a single AI framework — capable of addressing all of them. That integration is what made a project of this magnitude possible.»
