Artificial intelligence in the food sector: a real disruptor or just another hype?
Artificial intelligence (AI) is having a huge impact on everyday life. Whilst some are slow to adopt the applications, recent advancements are delivering great benefits to many industries. One such sector is food processing. In the most part, the food sector is a very high volume, low margin industry. This means finding a way to gain efficiency increases can make a difference between a business turning around a profit or a loss.
What is artificial intelligence anyway?
When talking about AI, many of us jump straight to sci-fi movies and think of Skynet creating Terminators capable of destroying the world. Arguably, in time, AI could develop the consciousness to be this intelligent (probably a discussion for another time) but right now, there is a somewhat softer definition. AI systems are any that demonstrate some behaviors associated with human intelligence. That usually consists of planning, learning, reasoning or perception.
In the world today, we mainly see narrow AI such as Alexa, Siri and Google Home that can take commands and learn from those. As well as this, recommendation platforms like Netflix, Spotify and Amazon would use a form of AI known as machine learning. In the food sector, there are several ways in which these narrow forms of AI can improve processes, safety, compliance, and margins. Below are short summaries of some of the key disruptors
Food sorting
In manufacturing and food processing factories every item is different. A plant might be sorting millions of carrots or apples per day into shapes, sizes, and colors. Towards the end of the 20th century, 90% of all the sorting was done manually. A lot of that is now automated reducing labor costs and increasing speed. It uses sensor and camera technology to detect anomalies or sort the food. For example, the machine is trained using data to know which potatoes to use for different applications e.g. make French fries or potato chips. It can quickly reduce food waste by optimizing the produce.
Supply chain
Emerging technology like Blockchain can monitor food from the farm all the way through to the supermarket shelves. Along the way, it is monitored for key attributes like temperature. If the temperature of the meat is too high or low, it breaks the chain and those involved in the supply will be alerted.
Other AI systems are tracking supply and demand of products to ensure effective supply chain management. Farmers can be told what to grow, when and how much of it they need.
Safety and compliance
Food hygiene rules such as Hazard Analysis and Critical Control Points (HACCP) which identifies, evaluates and controls safety hazards are now using AI to make them more efficient. This uses machine learning applications to classify hazards and create data alerts when they are suspected, like how Blockchain operates in the supply chain above. A manual HACCP process is time-consuming and offers limited traceability. Automated systems continuously collect data for effective auditing and root cause analysis.
Some restaurants use image recognition determine when areas do not meet compliance regulations. For example, cameras in the kitchen make sure employees where hair protection when required.
Some research has also gone into cleaning equipment using AI. Ultrasonic sensors and optical imaging can measure food residue and debris on a piece of equipment and initialize the optimal cleaning process. This creates a healthier and more affordable food industry.
Vegetation health-checks
SeeTree uses drones and machine learning to create health profiles for individual trees in giant orchards. This ensures greater yields using far smaller footprints.
Food flavoring
IBM has developed an AI system that could change the way we eat. The system can use data to predict how new flavor combinations will impact the consumer palette and the likely effect on their preferences. For example, it can work out if two ingredients will taste and smell like something people will want to purchase.
AI-Powered nutrition apps
Apps like Foodvisor are under development. This app takes a picture of a plate of food and can provide a provisional nutrition report to the consumer. It is a great way to instill personalized coaching so that users can keep track of their eating habits without manually writing everything they do. This could be quite revolutionary when it comes to weight loss. As users upload more images, the app continues to learn and will only become more powerful as it gets more subscribers.
Other apps like Nutrino Health use similar systems to combat diseases such as diabetes. It monitors glucose levels and reaction to certain foods.
Recommendation systems
AI in the food sector is now helping companies decide what the customer wants. For example, Coca-Cola has installed customizable drink fountains in several places where customers can select from a variety of options. This has created a mass of consumer data for Coca-Cola to analyze and recommend what products should be developed next.
What’s next?
AI can improve the production process and products in the food sector. The systems and technology have become a viable alternative for human expertise. AI can carry out tasks in seconds that would have previously been highly complex for humans to complete like food quality assessments. Labor costs and waste will reduce with those advancements.
Beyond this, using data for applications like machine learning has a wider impact in that it can help farmers manage crop growth, understand precise customer demand and predict what new products will work before anyone has even tried them.
The question being asked is how quickly AI will infiltrate its way through the food sector. Not if this will happen.
