Artificial intelligence finds patterns no human could spot
In a Málaga laboratory, scientists are teaching artificial intelligence to see what centuries of craft have missed — the microscopic damage hidden inside cured ham. A three-year collaboration between biomedical researchers and a 128-year-old Segovian producer is applying the tools of human medicine to one of Spain's oldest culinary traditions, seeking to refine the bleeding process that shapes the final quality of ibérico and serrano hams. The project is not merely about aesthetics; it is about what happens when ancient knowledge meets the patience of machines that never tire of looking.
- Current ham-bleeding methods, largely unchanged by modern food science, leave internal bruising and discoloration that diminish the product's appearance and potentially its flavor.
- Researchers at IBIMA are adapting biomedical biopsy techniques — designed for human tissue — to analyze ham samples far larger and more structurally complex than anything their methods were built for.
- AI algorithms are being trained to detect microscopic tissue damage patterns invisible to the human eye, automating the analysis of thousands of samples with speed and consistency no team of scientists could match alone.
- The three-year VACU4HAM project has already confirmed Monte Nevado's hypothesis: the current bleeding process is causing internal damage — and the data now points toward a better path.
- A prototype bleeding machine is under development that could be patented, potentially setting a new industry standard and reshaping how Spain's entire ham sector handles this foundational step.
In a Málaga laboratory, scientists are training artificial intelligence to solve a problem that has troubled ham makers for generations: how to drain blood from a cured ham without damaging the meat within. The project, VACU4HAM, unites Monte Nevado — a Segovian ham producer with 128 years of history — with researchers at the Biomedical Research Institute of Málaga, in an unlikely meeting of ancient craft and modern computation.
The bleeding step, performed after slaughter, relies on techniques that leave their mark: small tears in the tissue, discoloration that spreads like a bruise through the meat. None of it threatens food safety, but in a product where appearance carries as much weight as flavor, the damage is real. Alejandro Domínguez, who leads IBIMA's histology unit, frames the goal simply — improve the technique while eliminating the defects the current process produces.
What sets the project apart is its method. The team is applying microscopic analysis tools developed for human biopsies to ham tissue — samples far larger and more complex than anything biomedical science typically handles. They examine how different bleeding approaches alter the meat's internal structure, then feed that data to AI systems trained to recognize patterns no human inspector could reliably spot. Adapting these tools required rebuilding their methods almost entirely, since aged, salted ham behaves nothing like fresh human tissue.
The project runs until May 2028. Over that time, the team will test different bleeding systems across different stages of production, tracking how each affects tissue integrity and final quality. The ambition extends beyond fewer blemishes — avoiding tissue damage may also preserve flavor compounds that current methods inadvertently destroy.
Monte Nevado arrived at IBIMA with a suspicion; the science has confirmed it. Innovation director Alejandro Olmos says the company expects to develop a patentable prototype bleeding machine — one that could redefine industry standards. The details remain undisclosed, but the direction is clear: artificial intelligence is beginning to reshape one of Spain's most storied culinary traditions, from the inside out.
In a laboratory in Málaga, scientists are training artificial intelligence to solve a problem that has plagued ham makers for centuries: how to bleed a cured ham without damaging the meat inside. The project, called VACU4HAM, is a partnership between Monte Nevado, a ham producer in Segovia founded 128 years ago, and researchers at the Biomedical Research Institute of Málaga (IBIMA). It represents an unusual collision between ancient craft and modern computation.
The current bleeding process—the step where blood is drained from the ham after slaughter—uses methods that are, by the standards of modern food science, outdated. These techniques leave marks. They create small tears in the tissue. They cause discoloration that spreads through the meat like a bruise. None of this affects whether the ham is safe to eat, but it does affect how it looks, and in a product where appearance matters as much as taste, that is a real problem. Alejandro Domínguez, who heads the histology unit at IBIMA, explains that the goal is straightforward: improve the bleeding technique for both ibérico and serrano hams while eliminating the tissue damage and aesthetic flaws that the current system produces.
What makes this project unusual is the method. The researchers are applying techniques developed for studying human tissue—the kind of microscopic analysis doctors use on biopsies—to ham. They slice samples several centimeters across, far larger and more complex than the tiny tissue samples taken from patients. They examine these samples under magnification, looking at how different bleeding methods affect the internal structure of the meat. Then they feed this data into artificial intelligence systems trained to recognize patterns invisible to the human eye.
The challenge has been substantial. Human tissue behaves one way; cured ham, which has been salted and dried over long periods, behaves another. Fresh tissue and aged tissue are not the same thing to analyze. The researchers had to essentially rebuild their methods from the ground up, adapting biomedical science to food production. Domínguez notes that the artificial intelligence allows them to identify very small patterns that would be difficult or impossible to spot manually, and to automate the analysis of thousands of samples quickly and reliably.
The project is three years long and currently in its first phase. Over the next two years, until May 2028, the team will test different bleeding systems and different stages of the process to see how each one affects the tissue and the final quality of the product. The real challenge, Domínguez says, is creating a new system that is more efficient, produces less contamination, and maintains or improves the quality and flavor of the ham. The goal is a finished product with fewer defects—no spots, no darkening inside the meat—and potentially even better flavor, since avoiding tissue damage means avoiding the breakdown of compounds that affect taste.
Monte Nevado came to IBIMA with a hypothesis: the current bleeding method was causing internal tissue damage. The scientific work has confirmed this. What the researchers are doing now is providing data that the company cannot see on its own, information that will guide them toward a better process. The company's innovation director, Alejandro Olmos, says they expect to develop a novel prototype for bleeding ham—a new machine that could be patented and potentially reshape how the entire industry does this work. The details of how it will function cannot yet be disclosed, but the implications are clear: artificial intelligence is beginning to remake one of Spain's oldest industries, one ham at a time.
Notable Quotes
The current system is outdated and produces certain tissue damage and aesthetic defects— Alejandro Domínguez, IBIMA histology unit director
We will develop a novel prototype for ham bleeding that could be susceptible to patent protection— Alejandro Olmos, Monte Nevado innovation director
The Hearth Conversation Another angle on the story
Why does it matter how you bleed a ham? Isn't the meat the same either way?
The meat is safe either way, yes. But the current method leaves marks—small tears, discoloration. It's like the difference between a bruise that fades and one that stays visible. In a product people pay premium prices for, that matters.
So this is really about appearance, not safety or taste?
Not just appearance. The tissue damage affects how the meat cures, which affects flavor. And the aesthetic defects—they're a sign that something went wrong internally. Fix the process, and you fix both.
Why did it take artificial intelligence to figure this out? Couldn't scientists see this under a microscope before?
They could see it, but not at scale. You can examine one sample carefully, or you can examine ten thousand samples and find patterns no human could spot across all of them. The AI finds what's common, what's different, what matters.
Is this really a revolution, or is it just incremental improvement?
It depends on what the new machine can do. If it eliminates the defects while keeping the flavor, it changes the economics of the industry. Every ham producer will want it. That's not incremental.
What happens if the prototype doesn't work?
Then they've learned something valuable about why the current system is the way it is. But the scientists seem confident. They've already confirmed what the company suspected. Now they're just optimizing.