Hello

Your subscription is almost coming to an end. Don’t miss out on the great content on Nation.Africa

Ready to continue your informative journey with us?

Hello

Your premium access has ended, but the best of Nation.Africa is still within reach. Renew now to unlock exclusive stories and in-depth features.

Reclaim your full access. Click below to renew.

Aflabox: The AI tool transforming maize safety in Kenya

Aflabox SRL CEO Luca Alinovi (left) briefs Cereal Millers Association CEO Paloma Fernandes (centre) and Mombasa Maize Millers Limited Quality Assurance Manager Peninah King’ori on how aflatoxin is detected during the Annual Technical Conference & Expo at Sarit Centre Expo Hall, Nairobi, on April 15, 2026.  

Photo credit: Francis Nderitu I Nation Media Group

What you need to know:

  • AI-powered innovation is helping detect aflatoxin in maize quickly, offering safer food solutions for Kenyan consumers.
  • Scientist harnesses artificial intelligence to combat aflatoxin contamination threatening food safety and public health in Kenya.

When Luca Alinovi received a transfer offer to his home country Italy, he chose to remain in Kenya. He had been serving as country director for the Food and Agriculture Organization and had witnessed firsthand the country’s food security challenges.

One of the key issues he hoped to address was aflatoxin contamination in cereals, which affects human health and causes losses for farmers. “I found it incredible that we had so much aflatoxin in Kenya and we had to burn a lot of maize,” he tells the Nation in an interview on the sidelines of the Cereal Millers Association Expo in Nairobi. “We could not solve that situation. I took a radical decision to stay behind and find a solution.”

He focused on maize, Kenya’s staple food. His initial idea was to create a company that decontaminates aflatoxin in maize, leading to the birth of Aflazero. Although the first two pilot phases were conducted in Kenya, progress was slow and did not fully solve the problem, he says.

“It was a long process and we found it so difficult. We had to have regular testing in a laboratory and we had to buy consumables which were often out of stock. It was such a difficult thing for something that would ideally be simple,” he says.

His team then turned to further research, seeking faster and more effective solutions to ensure Kenyans consume aflatoxin-free food. “I had a super smart team of young Kenyans who came up with a prototype and we tried it and saw that it worked,” he says.

The prototype used artificial intelligence (AI) to scan and detect aflatoxin in maize, helping users separate safe grain from contaminated batches. The innovation won the Best Agro-tech Idea award at the Maker Faire Rome in 2023.

Last year, the prototype was fully developed into a product known as Aflabox, which was showcased at this year’s Cereal Millers Association Expo. Luca explains that Aflabox uses artificial intelligence and image processing to detect aflatoxin. Traditionally, identifying contaminated maize requires complex laboratory testing and chemical consumables.

Aflabox, however, captures images of fluorescence produced by aflatoxins under ultraviolet light. “It doesn't require any capacity from the operator. You just have to push a button after putting the sample of maize inside the drawer, close it and then start the test,” he says.

Once the system captures images of the maize sample, it sends the data to a cloud server, where the AI determines whether toxins are present. Results are then sent back instantly and displayed on the device. The entire process takes between one minute and 90 seconds.

The portable device can be recharged and costs about $5,000 (Sh646,000), with an annual fee of $100 for cloud data storage. Luca notes that the system allows users to track aflatoxin levels in their grain over time, improving monitoring and traceability.

Quality analysis 

Beyond detecting aflatoxin, the tool also analyses grain quality using white light imaging. It can identify defects such as broken grains, insect damage, and discolouration. “Whatever defect can be in the sample can be easily recognised and accounted for. That ends up being transparent in grading and makes sure that you have a full understanding of the grading,” Luca says.

While the technology initially focused on maize, the team is expanding it to other cereals, including sorghum and nuts. “We have started to train the AI on sorghum and other grains depending on what our clients need. It normally takes a few months to train the AI on a specific product, or on a specific toxin, because it's not only aflatoxin. There are multiple mycotoxins, which can be dangerous, and we have to train those as well,” he says.

Aflatoxin, produced by fungi in poorly stored grain, is invisible, tasteless, and resistant to milling. Once it enters the food chain, it cannot be removed. Across Africa, the toxin contaminates an estimated 40 per cent of grain supply each season. It remains a major public health concern in Kenya, particularly in maize and groundnuts, with exposure linked to liver cancer, stunting in children, and acute poisoning in high-risk regions such as Eastern Kenya.

Recent surveillance from the Kenya Bureau of Standards shows that a significant proportion of maize and maize products in local markets still exceed the regulatory limit of 10 parts per billion, especially in informal markets where enforcement is limited.

Alice Kemunto from Consumer Networks says cereal producers must educate consumers about aflatoxin risks. She also called on the government to enforce stricter regulations to prevent contaminated products from reaching the market. “We see those being flagged by the government and we still find them on the shelves. We should have proper working relationships between industry players and the government,” she says.