Tyson Foods Boosts Efficiency with Computer Vision and Machine Learning from AWS

Tyson Foods Inc. (Tyson Foods) processes millions of pounds of beef, pork, chicken, and prepared foods each week in its facilities worldwide and strives to achieve operational excellence throughout the production process. With large-scale operations, every manual step adds up, creating potential bottlenecks in the production process. That’s why the emerging technology team at Tyson Foods sought opportunities to automate processes and improve common time-consuming or error-prone tasks with computer vision (CV) solutions powered by machine learning (ML).

Investing in Computer Vision Technology

Tyson Foods has the capacity to process 40 million chickens per week, so it’s important for the facilities to run efficiently and have accurate inventory measurements to fulfill customer orders. Manual processes for keeping production on track, like inventory counting and machine inspections, cost valuable employee time and don’t offer near real-time insights at scale. To help automate these manual inspection processes, Tyson Foods wanted to incorporate CV, a process that involves capturing, processing, and analyzing images and videos so that machines can extract meaningful, contextual information from the physical world.

The company’s interest in pursuing CV technology to automate manual processes came in 2018 when it began its cloud migration to Amazon Web Services (AWS). During the migration process, the leader of the emerging technology team at Tyson Foods visited the Amazon Go flagship store in Seattle, Washington. The store’s use of cameras to automate the checkout and retail experience inspired Tyson Foods to use CV in its own facilities to reduce the need for manual inspections. The company successfully developed a CV solution to augment manual inspection processes but found that implementing CV can be complex and difficult to scale. The Tyson Foods team wanted to explore incorporating ML, which can help automate time-consuming processes, into their CV solutions to increase efficiency and decrease complexity even further. In 2021, Tyson Foods approached AWS for support with implementing CV solutions powered by ML to address inventory management and product carrier failure identification.

Simplifying Inventory Tracking for Near Real-Time Insights

Tyson Foods wanted to develop an automated solution for inventory tracking because manual techniques for counting chicken trays that pass quality assurance measures aren’t accurate enough. Alternate strategies like monitoring the hourly total weight of production per rack don’t provide data right away, so team members aren’t able to address issues promptly. Tyson Foods collaborated with the Amazon Machine Learning Solutions Lab (Amazon ML Solutions Lab), which pairs an organization’s team with ML experts, to build and train an object detection model. To do this, they used Amazon SageMaker, fully managed infrastructure, tools, and workflows to build, train, and deploy ML models for any use case. This model automatically detects and counts chicken trays on video streams from production lines as employees load them onto carts. Using AWS Panorama, a collection of ML devices and a software development kit that brings CV to on-premises cameras, the company was able to deploy this model at the edge to analyze video in milliseconds with low latency. With this CV solution, poultry production supervisors receive near real-time insights into production quantity and can avoid both underproduction and overproduction during the shift.

“These solutions help us use exactly what we need by understanding the true

and optimizing inventory so that we can effectively plan and reduce waste.”

Barret Miller, senior manager of the emerging technology team at Tyson Foods

Automating Maintenance Inspections

Tyson Foods also aimed to automate the task of recurring inspection of the plastic pins used to hold product carriers in place in its poultry production facilities. This inspection process required attention to detail and valuable operator time. Because a pin falling out could cause safety issues or unplanned downtime, employees spent an estimated one hour per shift inspecting nearly 8,000 pins per line.

For this solution, Tyson Foods turned to Amazon Lookout for Vision, an ML service that uses CV to spot product defects in objects at scale. Using Amazon Lookout for Vision, the company created a custom ML model to analyze images and detect anomalies without needing ML expertise. Tyson Foods deployed this model at the edge on an AWS Panorama Appliance, which organizations can use to connect cameras and process multiple CV applications on multiple video streams simultaneously. By deploying the model at the edge, Tyson Foods employees are notified right away that a product carrier needs maintenance when the model identifies anomalies. With this solution, team members no longer need to spend time inspecting product carriers, which can save the company 15,000 hours of skilled labor annually in a single facility.

Continuing to Innovate to Reduce Manual Intervention

The emerging technology team at Tyson Foods continues to develop expertise with every new CV solution, addressing production needs faster and automating more of the business. Tyson Foods plans to continue using AWS services to innovate and develop additional CV solutions that are powered by ML to reduce the need for manual intervention, improve efficiency, and optimize processes.

Learn more about industrial ML services from AWS, such as those used by Tyson Foods, at aws.amazon.com/industrial/machine-learning.