Tyson Foods taps into data at the edge to modernize protein production

From analyzing logistics networks to tracking chickens with computer vision, the 85 year-old food processing giant is finding new ways to stay competitive in a fast-changing industry.
Written by Stephanie Condon, Senior Writer

When Scott Spradley joined Tyson Foods as CTO in 2017, the multinational corporation was ready for a major technology and business overhaul. With nearly $40 billion in annual sales that year and more than 120,000 employees, the business sat comfortably within the Fortune 500. It produced around 20 percent of the beef, pork, and chicken consumed in the US. But it had effectively no cloud computing presence to speak of. 

"When I walked in the door, it was very much an on-prem set of environments, very stable - even mainframe," Spradley told ZDNet.

Coming from HPE where he served as CIO, it was clear to Spradley that the food industry wasn't exactly moving at Silicon Valley speeds. But Tyson's leaders, he said, were "hell-bent on getting Tyson into the modern age of technology." 

The company kicked off a "multi-threaded transformation approach" built on a hybrid cloud strategy. Leveraging AI, machine learning and predictive analytics, Spradley said Tyson could create more efficient operations and keep closer tabs on consumer preferences. It could even start taking advantage of cutting-edge technologies like computer vision to improve its chicken and pork production. 

Ultimately, Spradley said, Tyson "should aspire to be a tech company that happens to deliver protein."

They started by implementing a whole new SAP stack, Spradley said, giving the company the ability to get richer analytics -- and a deeper look into its supply chain. Spradley said his team asked, "What do we need to do to predict what's going to be consumed and at what volumes... to identify generational trends around food consumption and how that affects the capacity we need?" 

That capability, Spradley explained, can have a significant impact in areas like poultry production, where Tyson is fully vertical -- from laying eggs to providing specific kinds of chicken products to fast food brands like Wendy's. 

Along with leveraging analytics, Tyson started improving its security posture and adopting cloud products like identity management tools from Okta. 

"We've done a lot of the blocking and tackling," Spradley said. 

With those components in place, Tyson was able to extend its network beyond corporate knowledge workers, Spradley said. This has allowed the company to pull data from the edge, enabling it to build truly predictive models and leverage more sophisticated technologies. 

"The edge can be a beef plant operation in South Dakota, where we're looking at worker data," Spradley said, "having computer vision watching the way people on the line are operating and giving us machine learning intelligence that can say, 'Here's a faster process that has a higher quality."

Tyson is also working towards using facial recognition on its herds. It's already using computer vision in some of its poultry houses. 

"It tells us exactly how many days a baby chick takes to get to the exact specification" of a customer order, Spradley explained. "We can also learn through that computer vision how much movement the bird should have to get the perfect chicken."  

Tyson leverages tools from cloud providers as much as possible, Spradley said, rather than building custom components. The company is using services from both AWS and Google Cloud Platform. 

"We're in a fortunate position these days," Spradley said, likening the competition between AWS and other cloud providers to the nuclear arms race of the 1980s. "AWS and Google are disrupting each other almost quarterly, creating new technology."

All told, Spradley's team has around 430 projects it's trying to accomplish this year, mostly around business transformation and refining processes. Transportation logistics, he said, offers "the sweetest spot for refinement."

"We're the biggest protein deliverer in the United States, so we have trucks all over the US all the time," he said. "And we're looking for ways to refine that to reduce the throughput time, from the idea of a product to getting it on the shelf or in a restaurant."

Currently, amid the COVID-19 pandemic, "one of the biggest goals is to make sure the nation gets all the food they need," Spradley said. "That's kind of our rallying cry right now as a company -- we're kind of counted on to make sure grocery stores are stocked."

Editorial standards