Input Stories

Acquiring an AI-optimized mill has helped boost U.S. Steel Corp.’s bottom line—and is giving the 121-year-old manufacturer “inspiration about what is possible” in its other facilities, according to The Wall Street Journal (subscription).
 
What’s happening: “Big River [steel plant in Osceola, Arkansas] uses advanced technology to make basic steel mill functions, such as the cooling of hot steel coils, more efficient. If the coils are too close to one another, they take longer to cool, which is why Big River’s machine-learning automated crane is so important.”

• “Big River also uses cameras to feed inputs into machine-learning algorithms that can detect defects in coil slabs or determine whether someone creates a safety hazard by getting too close to certain machines.”

Increased profitability: The purchase of the Arkansas plant has already paid for itself and led U.S. Steel to its highest-ever first- and second-quarter earnings this year.

Not easily replicable: But Big River’s successful algorithms can’t just “plug and play” into other U.S. Steel mills. Each facility must construct and train its own models based on the unique environment.

• “That means the transformation of older mills involves the installation of data-generating devices such as sensors and cameras onto existing equipment, according to U.S. Steel Chief Information Officer Steve Bugajski.”

Other challenges: Upgrading older steel mills, many of which have equipment from the 1950s and ’60s, comes with obstacles. Internet bandwidth is poor in buildings with so much steel and concrete, and older machinery can’t be retrofitted with new, internal safety sensors.
 
Change in action: At the manufacturer’s 10-year-old Gary Works mill in Gary, Indiana, several major Big River–inspired updates are being made.

• “An area called the hot strip mill, where slabs of steel are converted into coils, can create bottlenecks in production. … Now a machine-learning system can analyze why certain slabs took longer than they should have.”
• Gary Works is also working to create a digital twin, or a live virtual representation of what some key equipment is doing at any given time, with the ultimate goal of predicting finish times and optimizing output.”

Learn more: Interested in the overall trends in manufacturers’ AI adoption? Check out this recent survey from the Manufacturing Leadership Council, the NAM’s digital transformation arm.