Food safety remains one of Nigeria’s most persistent yet often overlooked public health challenges. From poor storage conditions and post-harvest losses to limited testing facilities and inconsistent inspection standards, ensuring that food reaching consumers is safe remains a complex task.
Kenneth Sambo believes artificial intelligence could become part of the solution.
In an exclusive interview with AIBase.ng reporter Ahmad Ibrahim, Sambo, an expert in artificial intelligence, shared his views on how AI, computer vision, and sensor technologies could help improve food safety monitoring across Nigeria’s food supply chain.
According to him, his interest in AI-driven food-safety systems emerged from concerns about the limitations of traditional inspection methods, which remain heavily dependent on human observation.
“The traditional method is what I personally see as an issue,” Sambo said. “Human inspection can be affected by fatigue, subjectivity, experience and the volume of food being inspected, which can lead to inconsistencies.”
He noted that food-safety inspections are often time-consuming and expensive, while some forms of spoilage or contamination can be difficult to identify with the naked eye. Laboratory testing can provide additional certainty, but such facilities are not always available to every food-processing operation.
One example that particularly influenced his thinking was the process of selecting fruits and other raw materials for food production.
“When fruits are being selected for juice production, the sorting may sometimes depend heavily on manual visual inspection,” he explained. “A fruit may appear healthy externally but have internal defects. Factors such as ripeness, storage conditions and intended use also matter.”
Those observations led him to question whether artificial intelligence could provide an additional layer of objective screening before food enters the production process.
“This made me ask whether AI could provide an additional layer of objective and consistent screening before the food enters production,” he said.
For Sambo, the issue extends beyond manufacturing efficiency. He argues that food safety is fundamentally a public-health concern.
Many consumers, particularly in environments where affordability and access to food are pressing challenges, may prioritise having food available over evaluating its nutritional quality or safety. Unsafe food, he noted, can contribute to illness and other health complications.
That is where he believes AI could offer meaningful support.
According to Sambo, advances in computer vision have created opportunities for machines to continuously analyse food images and identify patterns associated with spoilage, contamination indicators, and physical defects at a speed and consistency that would be difficult to achieve through manual inspection alone.
The system he is developing combines computer vision, machine learning and potentially sensor technologies. Images and data collected from food samples are used to train an AI model to recognise specific conditions. When a new sample is presented, the system analyses its characteristics and determines whether it falls into categories such as acceptable, spoiled or potentially defective.
Additional sensors could also provide information beyond what images can reveal.
“Images alone cannot tell you everything,” Sambo explained. “Sensors could provide additional information such as temperature, humidity and, where appropriate, measurements related to microbial or chemical hazards.”
One of the biggest advantages of computer vision, he said, is its ability to identify subtle visual patterns that may not always be consistently detected during manual inspections.
He pointed to indicators such as discolouration, bruising, cracks, surface abnormalities, and mould-like growth as conditions that AI systems could potentially detect at scale.
To illustrate the challenge, Sambo asked readers to imagine several tonnes of oranges arriving at a processing facility.
“Manually sorting that volume requires substantial labour and time, and consistency can become difficult,” he said. “A computer-vision system could potentially inspect the fruits continuously on a conveyor and flag those showing characteristics associated with defects or spoilage.”
Such a system could make the initial screening process significantly faster and more consistent while reducing manual effort.
However, Sambo was careful not to overstate the capabilities of current AI systems.
“I would not claim that computer vision can detect every form of contamination,” he said.
Some food-safety risks, including certain bacteria and chemical residues, cannot be identified through visual inspection and would still require laboratory analysis or specialised sensing technologies.
As a result, he does not view AI as a replacement for traditional food-safety science but rather as a complementary tool.
The project itself remains in the research stage. Sambo said he is not yet prepared to claim a final commercial accuracy figure because the system still requires extensive validation.
“At this stage, I would describe it as a research and prototype project,” he said.
Performance would ultimately be measured using established machine-learning evaluation methods, including accuracy, precision, recall and other validation metrics tested against representative datasets.
Sambo believes the need for innovation is particularly pressing in Nigeria, where food safety challenges remain widespread.
He pointed to poor storage conditions, inadequate cold-chain infrastructure, food contamination, improper handling practices, limited testing facilities and significant post-harvest losses as some of the sector’s most pressing problems.
Nigeria’s large informal food sector adds another layer of complexity.
“Ensuring consistent food-safety practices across farms, markets, transportation and processing facilities can be difficult,” he said.
Despite those challenges, Sambo believes AI-powered inspection systems could be adapted for Nigerian conditions if affordability and accessibility are prioritised.
He envisions solutions built around smartphones, low-cost cameras and portable sensors that can operate even in environments with limited internet connectivity.
“The system should be affordable, portable, easy to operate and capable of working with limited internet connectivity,” he said.
Equally important, he added, is ensuring that AI models are trained on Nigerian food samples and local environmental conditions rather than relying solely on datasets collected elsewhere.
Like many emerging technologies, however, adoption will depend on more than technical performance.
Sambo identified cost, infrastructure limitations, technical expertise and trust as some of the biggest barriers businesses may face when considering AI-based food-safety tools.
“There is also the issue of trust,” he said. “Businesses and regulators would need evidence that the technology works reliably before using it for important food-safety decisions.”
Potential users of such systems could include food-processing companies, agricultural businesses, supermarkets, laboratories, quality-control departments, regulatory agencies and large food distributors.
Related
- Google Africa AI Investment: $225 Million to Build an AI-Ready Data Future
- What Is AI? A Simple Explanation for Nigerians
Although he sees commercial potential in the long term, Sambo says his immediate focus remains on research and validation rather than launching a product.
“I want to establish that the system works reliably before thinking seriously about taking it to market,” he said.
The next phase of the project will focus on collecting larger, more diverse datasets of Nigerian food products, improving the machine-learning models, and testing them under real-world conditions.
He also hopes to integrate computer vision with additional sensor technologies to create a system capable of detecting a broader range of food-safety concerns.
Looking ahead, Sambo believes AI could fundamentally change how food safety is monitored in Nigeria over the next five years.
Instead of relying entirely on periodic inspections, AI could support continuous monitoring across the food supply chain, from production and storage to processing and distribution.
Yet he remains clear that technology should support experts rather than replace them.
“I don’t see AI replacing food scientists, inspectors or laboratory testing,” he told AIBase.ng. “I see AI as a tool that helps them make faster and better informed decisions.”
For Sambo, the future of food safety lies not in removing human expertise from the process but in augmenting it with technologies capable of processing information at a scale and speed that was previously impossible.
“AI should therefore be seen not as a replacement for professional expertise,” he said, “but as an additional tool for creating a safer and healthier food environment.”
