Using AI To Increase Food Quality Posted on : May 08 - 2021

When talking about food quality, AI isn’t usually the first thing that comes to mind. But by integrating AI into the food manufacturing process, companies can maximize efficiency in quality control. According to Mordor Intelligence, artificial intelligence in the food and beverage market was valued at $3.07 billion in 2020 and is expected to reach $29.94 billion by 2026 at a CAGR of over 45.77% during the forecast period of 2021 - 2026.

One of the pillars of food quality is the safety of the food. Reducing the presence of pathogens and detecting toxins in food production is a key part of AI. Japanese company Fujitsu has developed an AI-based model which is used to monitor hand washing in kitchens with six-step hand washing regulations set by the Japanese health ministry. Fujitsu’s model builds on its existing behavioral analytics capabilities, which can already recognize a variety of subtle and complex human movements using deep learning techniques without relying on large amounts of training data. The technology captures images of complex handwashing movements as a combination of hand shape and repetitive rubbing motions, using two deep learning engines: hand shape recognition and motion recognition. With a handwash video dataset comprising about 2,000 variations of people, camera positions, and soap types, Fujitsu said its technology was able to detect the six-step hand washing process with an accuracy rate of over 95%. Using AI in this part will definitely reduce the need for visual checks where food safety must be increased, especially in the COVID-19 process.

Another pillar is the processing part. In this part, AI can maximize the output and reduce waste by replacing people on the line whose only jobs are to distinguish identify items unsuitable for processing. Electric noses are the ones to be used in processing parts and can be used as replacements to human noses to distinguish various foods’ odors and aromas using AI sensors. Research about using e-noses to detect the quality of olive oil is made by Hindawi. The research was held in Balikesir in Turkey, which is in the Mediterranean region with the majority of the olive trees (805 million) in the world. In this research, 12 different olive oil classes type has been constituted for both training and testing sets without mixing themselves. E-nose smells the odor of each olive oil samples’ aroma. It consists of 32 odor sensors. Then, these odor data are normalized by using the Z-transformation method after digitized. Finally, the types of olive oils are characterized by a machine learning algorithm using 32 inputs normalized data. This process automatically gives a quick quality check for olive oils. One of the biggest producers of this region is Ozgun Zeytincilik in the Balikesir area. “Adulteration with cheaper or lower quality oils have a huge impact on quality. Most frequent adulterations in olive oils can be seen with sunflower oil, maize oil, coconut oil, and hazelnut oil. So, using e-nose in this processing part is a game-changer in terms of quality checks,” said Halil Sucu, founder of Ozgun Zeytincilik.  View More