Artificial intelligence and robotics are taking the food industry by storm

The growth of a chicken is unpredictable. However, the Compass platform offers 14 days of weight predictability, so you know what day and time the animal will reach its target weight for processors and rotisseries. (Picture: courtesy)

FOOD PROCESSING. Artificial intelligence (AI) and robotics are becoming increasingly important in the food industry, from production to marketing, processing and distribution. Quebec is no exception to this trend.

When it was founded in 1999, Intelia manufactured electronic components. She began to specialize in agriculture in the mid-2000s, particularly in the production of meat and dairy products. It then supplies climatic controllers. “Around 2015, we noticed that producers were experiencing major challenges due to the complexity of production parameters,” explains Caroline Forest, vice-president of sales and marketing for the company. Concerns about animal welfare and new rules multiplied the elements to be managed. »

The Joliette company then decided to work on an AI-based data analysis tool, which would become the Compass platform. “We offer an integrated solution that includes sensor installation, data collection and analysis, and predictive models,” emphasizes Caroline Forest. A mobile application allows producers to have real-time access to a wealth of data. The models are continuously updated. They also receive alerts or notifications that can lead them to take action.

Caroline Forest gives the example of chickens that are produced for slaughterhouses that supply processors or rotisseries. The latter impose very precise requirements in relation to the size of the birds they buy, since they must supply very homogeneous products to their customers. Producers and packers earn bonuses when they ship poultry with the right characteristics. But a chicken is alive, and its growth keeps a dose of unpredictability. The Compass platform offers 14 days of predictability on the weight, to know on what day and at what time the animal will have reached its target weight.

Information from the platform can be shared in real time with processors, helping to optimize actions across the chain. “The objective is not to provide more data, but to help the actors of the chain to see more clearly in the large quantity of data which is already produced”, affirms Caroline Forest.

hurry up

At Cégep de Lévis, the Center for Industrial Robotics and Vision (CRVI) has also been focusing on agriculture and food processing for several years. “AI is a very relevant tool for areas that produce large amounts of data, which is really the case in all sectors of the food chain, from production to in-store sales”, underlines the director. General Yves Dessureault.

In food production and processing, AI is increasingly going hand in hand with robotization. For example, the CRVI has developed a broccoli picking robot in collaboration with the National Optics Institute (INO), on behalf of Lapalme Agtech. The system is paired with a tractor. It detects, positions and assesses broccoli using AI, to harvest only those that meet the grower’s size and maturity requirements. Such a tool will make it possible to increase the frequency of harvests, which is currently decreasing due to the lack of manpower.

AI and robotization are also playing an increasing role in quality and safety control. This is a major challenge for the food industry. There are an average of 200 recall incidents per year between 2015 and 2020, according to Agro Québec. Undeclared allergens (often because they are not detected) are regularly the cause.

“The democratization of certain technologies, such as hyperspectral imaging and artificial intelligence, makes it possible to create much more efficient systems for detecting diseases or singularities”, assures Yves Dessureault. Hyperspectral imaging relies on a much more precise molecular signature than traditional visualization provides.

Yves Dessureault gives the example of sugar and salt, which cannot be distinguished with a traditional system, but which have a very different molecular signature which allows them to be identified with hyperspectral imaging. These new systems therefore contribute to the detection of diseases and pathogens in food, but also to the monitoring of other data, such as the water content of a plant. This can optimize growers’ use of water.

In Quebec, only 32% of agri-food businesses have automated more than half of their systems so far, according to Agro Quebec. “With the labor challenges we are experiencing, the goal of increasing our food self-sufficiency can only be achieved through automation,” says Yves Dessureault.

Similar Posts

Leave a Reply

Your email address will not be published. Required fields are marked *