Labour challenges and increased demands for automation and traceability throughout the supply chain are driving implementation of technologies that automate various farming tasks and lessen the reliance on a human workforce.
Artificial intelligence can help make those robots even smarter and automate tasks — like harvesting produce for example — that previously could only be done by humans. It is also essential to helping analyze and interpret the vast reams of data constantly being collected on farms and across the entire agri-food supply chain so it can be used for more informed decision-making.
Why it matters: Artificial intelligence in agriculture overall is projected to grow at an annual rate of 25.5 per cent between 2020 and 2026, according to data from Research and Markets, with the artificial intelligence-as-a-service segment predicted to see the biggest growth.
Quebec has evolved into Canada’s artificial intelligence powerhouse. It’s home to the Montreal Institute for Learning Algorithms, the world’s largest academic research centre in deep learning, and the Institute for Data Valorization, for example, as well as the federally funded Scale AI Supercluster.
Scale AI was one of five superclusters that received a combined $950 million over five years in 2018, and includes a wide range of businesses, including some from the agri-food sector, like Agropur and Lallemand. Its main goal is to build intelligent supply chains that will make it easier and faster predict when and where products are needed and then meeting that need.
“We are proud to have over 1,300 companies as part of our active ecosystem,” said Scale AI CEO Julien Billot in French remarks provided at the 2020 virtual supercluster showcase last fall. “We support over 28 accelerators and incubators across the country and those incubators themselves will support more than 150 artificial intelligence start-ups each year.”
But Quebec’s artificial intelligence ecosystem is also encouraging innovations more directly linked to agriculture. A couple were presented at the recent online conference hosted by Glacier FarmMedia agricultural publication Le Bulletin des Agriculteurs.
Smart Counting automates the repetitive and time-consuming task of counting pigs. Developed by Conception Ro-Main of Saint-Lambert-de-Lauzon, the goal is to reduce on-farm labour requirements, as well as limiting stress on the animals.
“Smart Counting comes from the real need of producers to count pigs. Everyone has to count animals every time they are moved from one spot to another and it’s an annoying task,” said Jacquelin Labrecque, Ro-Main’s director of research and development in French during his presentation.
Automated counting also increases count accuracy. According to Labrecque, Smart Counting can provide accuracy to 99.9 per cent, whereas manual counting rate errors are estimated at five pigs out of a thousand.
The system consists of an overhead corridor camera that follows each pig individually as it moves through the camera’s field of vision. The captured video is sent to a server where it is analyzed by specially developed algorithms that are trained to count pigs of all sizes and colours. A mobile dashboard lets users track and manage their data in real-time, as well as generate reports, graphs and statistics.
Another intelligent management system comes from Agrimesh Technologies in St-Hyacinthe. It lets producers fully automate and control the interior environment of their livestock facilities, including ventilation, lighting or carbon dioxide concentration.
Sensors monitor barn conditions and artificial intelligence analyzes the data to calculate the best energy strategy to keep the inside target temperature consistent at set level, regardless of how the outside temperature might increase or decrease. The system then automatically adjusts itself based on the algorithm’s recommendations.
Louis Landry of Ventec, which distributes the wireless system in Quebec, likens it to having an additional employee in the barn who can make independent decisions to ensure consistent and optimal conditions for livestock or poultry.
“It’s more than just a room controller system, it’s a smart management system that is constantly making decisions and managing independently - when you are moving livestock, for example, it knows when to turn on the ventilation system automatically,” he said. “Intelligent management is a new way of functioning, a new way of thinking that we also see increasingly in other sectors.”