From farm to fork: AI shows the way towards safe and sustainable food production

Contents

• Food Safety Risks along the Supply Chain
• One Health - for People, Animals and the Environment
• Changing the path from Predictive Maintenance to Predictive Quality
• AI-enabled Predictive Quality
• Examples of AI-enabled Predictive Quality for Food Safety
• Safe and Sustainable Food Production

Food Safety Risks along the Supply Chain

The modern consumer holds food safety among their top 3 purchasing criteria, and is acutely concerned about safety risks ranging from microbiological to chemical. Many are careful to avoid allergens. Typical food production processes already mitigate such risks with strong quality control checks locally. However, new risks are constantly emerging across the supply chain, including mycotoxins, viruses, as well as increased fraud and adulterations. Further sources of upstream risks include greenhouse gases emissions, domestic and industrial waste, chemical pollution and pesticides. Moreover, current quality control practices are often inefficient, and complete quality measures along the entire supply chain are unavailable.

One Health - for People, Animals and the Environment

Additionally, in the field of public health, a similar thought has been forming - that in order to protect the health of human populations, it is essential to preserve the health of animal populations and ecosystems. The spread of zoonotic diseases such as Ebola and Covid-19 in recent times have made it clear that public health challenges need to be tackled as a problem of 'One Health' with an interdisciplinary approach that includes animal health and environmental health.

Consequently, the idea of supply chain quality management (SCQM) has emerged to combine supply chain management and quality management, which shifts the thinking of quality from an internal organisational view to a supply chain-wide perspective. Within the food industry this means recognizing that every step has an impact on the one after - monitoring and passing data to the next step means better health outcomes overall for humans, animal and the environment.

Changing the path from Predictive Maintenance to Predictive Quality

Borrowing from predictive maintenance being adopted by the manufacturing industry, the idea of SCQM implies that food manufacturing organizations now need to shift their thinking towards predictive quality. Traditional manufacturing leaders are highly familiar with predictive maintenance, where the goal is to minimize unplanned downtime by anticipating equipment breakdown. Similarly, Predictive Quality aims at anticipating and avoiding product quality deviations, in order to keep the focus constantly on product safety. For food production companies, predictive quality means not just safer food, but also cost savings and reduced waste, not to mention ensuring compliance as well as a stronger brand reputation.

AI-enabled Predictive Quality

With all the challenges listed above, it is humanly difficult to ensure predictive quality without having a clear data strategy across the food supply chain and enlisting the assistance of AI. Currently available cutting-edge AI can easily detect contaminants, allergens, and pesticides in food raw materials and processed food. Furthermore, AI can detect weak signals, much earlier in the supply chains, which can give a preliminary indication of trends or changes. This allows a proactive approach towards prevention and resilience in food safety, leading to reduced risks and greater opportunities to ensure quality. Needless to say, AI is a game-changer for not just food production companies but also distributors and regulators.

Examples of AI-enabled Predictive Quality for Food Safety

Let's take the example of detecting chemical contaminants in cheese production. Typically, trace amounts of chemical contaminants are present in milk, but impossible to detect and measure using traditional methods. However, even such trace amounts of pollutants concentrate in the finished cheese. By the time these are detected at the end of the production line, it can cause significant wastage and losses. With the Fujitsu Chromatography-on-Demand SaaS solution, high-resolution chromatography data is parsed by state-of-the-art AI models which are able to detect chemical contaminants at trace levels directly in the milk.

Let's take another example - detecting Listeria in factories. By the time typical microbiological analyses detect a Listeria crisis, it would have been well underway, resulting in a partial or total production shutdown. Alternatively, using routine microbiological analyses and contextual data including temperate and humidity, Fujitsu AI may detect even weak signals in fluctuations of bacterial populations to detect early warning signs. Similar processes can be ensured for early detection of E.Coli as well.

Safe and Sustainable Food Production

There is no doubt that AI is becoming a game changer for both safe as well as sustainable food production across the entire supply chain. As outlined in our AI Strategy, Fujitsu believes AI will transform as our primary assistant for most day-to-day enterprise activities. Ensuring food safety across the food supply chain will be no exception to this AI-enabled transformation. With our focus on Sustainable Manufacturing and Healthy Living, Fujitsu is well poised to act as your Sustainability Transformation partner.

For a brief introduction into our cutting-edge AI-enabled predictive quality for food and drug safety, check out this video: https://www.youtube.com/watch?v=c9L2wowD48s

A good place to explore further would be to get hands-on access to Fujitsu Kozuchi, a cloud-based platform, and try out our AI toolkits including Kozuchi for Vision, AutoML and Generative AI engines, for rapid development, testing and implementation. www.fujitsu.com/global/kozuchi.