Tyson Foods plans to implement cameras, machine-learning algorithms and edge computing at all seven of their “fresh tray” chicken facilities by the end of the year to help streamline inventory management and improve efficiency.
The company – which has invested more than $215 million in robotics and other automation technologies in the last five years – uses computer vision and artificial intelligence (AI) at three of their chicken plants.
“When you think of a consumer packaged goods (CPG) company, technology isn’t always the first thing that comes to mind. But top to bottom, technology is everywhere at Tyson,” Derek Burleson, Tyson’s Public Relations Manager, explained.
“There are unique challenges facing food companies like Tyson Foods when it comes to embracing AI and robotics — e.g., having to navigate cold, wet cold storage environments — so often integrating these technologies takes more time and research in the food industry.”
Moving past manual systems
In the past, chicken inventory was tracked manually. Workers counted meat packages wheeled in on carts, entering the data by hand. The process was prone to costly errors and could result in too much chicken being processed and consequently spoiled.
On the other hand, computer vision automates the tracking system, allowing Tyson to better control inventory and ensure the freshness of its chicken. A form of digital technology, computer vision uses cameras and computers to identify and validate the quantity of products that are processed in large quantities.
At Tyson, cameras identify the type of product, such as a package of chicken thighs, while an automated scale records the weight of each batch.
“Automating manual processes in poultry plants to move and track inventory in cold storage has led to precision going up by double digits with respect to cold storage inventory accuracy,” Burleson said.
Interest in computer vision has grown in the last few years as installation costs have dropped. Systems have become more affordable and the industry has seen improvements to both machine-learning algorithms and data-management capabilities.
Like what you just read? Sign up now for free to receive the Poultry Future Newsletter.
