SAP’s Industry Center in Delco uses artificial intelligence to streamline manufacturing

When Hurricane Helene damaged western North Carolina in September, filling medical facilities with people who couldn’t be cared for at home, Jim Polk, a tech executive at Exela Pharmaceutical Sciences in Lenoir, N.C., got a call from the Atrium Health hospital group in flooded Asheville. Could his company make intravenous saline bags for the expected crush of patients?

“That could take six months,” Polk thought, going over in his mind the familiar steps of meeting government and hospital requirements, hiring, lining up machinery and suppliers, writing and sharing reports, and weeks of factory trial and error to get production right.

But Exela had upgraded its manufacturing technology — and not just on the floor of its 600-person plant in Lenoir. Exela had also installed new software to connect factory production data to materials-sourcing, labor-supply, government-compliance, market demand, shipping, pricing, and other functions, using artificial intelligence applications to automatically pinpoint and fix problems.

This use of self-improving AI software to rapidly boost efficiency was dubbed a “Fourth Industrial Revolution” in a widely read 2011 book by Swiss engineer Klaus Schwab. He projected AI-enhanced manufacturing would transform factories as thoroughly as earlier steam power, mass production, and computer-control technologies. The Helene crisis gave Exela a chance to test these “Industry 4.0″ software applications on a new product line.

In less than three weeks, Exela was producing IV bags ready to ship.

“It was beyond miraculous,” Polk said. “We bring up a line now without adding heads. We don’t need to make and copy and retain all those documents. We are saving many weeks. We can immediately rock and roll.”

Polk and his colleague Don Overton, Exela’s finance chief, had seen demonstrations of what their new software could do — designing, mixing, bottling, and shipping chemicals, drugs, and soft drinks, among other sample products, as they provided a few directions, and watched the machines kick in.

They had seen how this could work on a late summer visit to the Industry 4.0 Center, a model factory complex at Exela’s main business software supplier, SAP, at its glass-walled U.S. headquarters in Newtown Square, Delaware County.

“It’s the future of manufacturing — purely digital, completely paperless,” and heavily automated, said Overton. “It’s seamless, and integrated in real time: Accounts payable is tied directly into purchasing, which is tied directly into warehousing, which is tied directly into production, which ties directly into shipping, and that ties into sales, and into HR.

“So when we moved product, there’s no delay. There’s no paper to reconcile. Records don’t accumulate. We call it Bye Bye Paper.”

The typical cost for SAP’s digital manufacturing execution system includes an annual subscription fee of around $100,000 — but also an outlay of at least $300,000 worth of new hardware to make it run. According to a June 2024 study by International Data Corp., users could expect 10% productivity savings — plus a tripling of their profitability over three years, IDC added.

SAP’s “copilot” AI application, Joule, takes questions from users across a manufacturer’s network. Asked about sources for popular materials, Joule will list prices, shipping times, reliability, labor and political factors, and other variables for suppliers around the world, varying results by producer and customer needs and locations.

Germany-based SAP employs around 3,000 at its U.S. headquarters in Newtown Square. The company also has Industry 4.0 Centers at a handful of other sites in the U.S., Europe, and Asia.

But with its suburban Philadelphia location near the center of the Northeastern United States, the Newtown Square center expanded last year from a smaller facility. It’s now about 8,000 square feet, roughly the size of a Super Wawa (the entire headquarters building is over 200,000 square feet.)

“I love that center. We have hosted customers, IT and business stakeholders there to explore [how to speed up] supply chains and make continuous improvements,” Asif Poonja, CEO of Fujitsu Americas, said in a phone interview from his office in Toronto.

Once a Japanese manufacturer, Fujitsu isa global manufacturing adviser to companies such as ABB, the Swedish generator and electrical equipment maker. Fujitsu has brought engineers and managers from ABB and other global companies to Newtown Square to demonstrate AI’s impact on factory automation and business processes.

The center, with its design and control centers, factory robots, and bottling and filling machines, makes software use more tangible: “Sitting on a Zoom call or on a laptop is not always effective” at engaging factory professionals, Poonja said. “To see these solutions in action” on actual machines designing, producing, and packaging chemicals, drugs, or beverages “is very powerful.”

Poonja says the new-generation manufacturing software also helps hire and train workers faster.

“It’s hard to get into factories these days. In Newtown Square you can go right on their [working model] factory floors and have discussions with their engineers. It’s amazing the details you can get into,” Poonja added.

SAP employs 112,000 worldwide. SAP shares have risen twice as fast as rival Oracle over the past five years, and also beaten its U.S.-based rival over the past one- and two-year periods.

Does all this automation destroy jobs? Exela said it’s been hiring even as it’s been automating production, because overall sales are up.

But automation has had a marked impact on next-generation factories. For example, GSK, the pharmaceutical giant that once employed thousands stamping pills at Philadelphia-area sites, plans to invest up to $800 million (including $21 million in state funds) in a factory near the Susquehanna in Marietta, Pa., that projects a modest 200 jobs.

SAP’s Industry 4.0 Center includes machines from a constellation of companies, many based in SAP’s homeland.

There are mixing machines from Beckoff, a real-time location systems from Kinexon, a storage-and-retrieval system from Gebhardt, and high-speed inspection machines from Krones, all based in Germany.

There are handheld and wearable scanners from Illinois-based Zebra. There are robotic carriers from a Danish company, Universal Robot, mounted with cameras from based-in-Germany Asentics; Mobile Industrial Robots from MiR, another Danish company; and a “hands-free” maintenance system using Germany-based Frontline’s TeamViewer software and Washington State-based RealWare’s Navigator headsets.

“With all the changes going on with the supply chain, all the disruptions” in recent years from COVID shutdowns, attacks on Red Sea shipping, and port labor strikes, “we wanted a facility where we could highlight our solutions and show how these applications work together to solve problems — instead of having folks use a lot of Power Points,” said Darcy MacClaren, SAP’s chief revenue officer for digital supply chain, during a visit to the center.

Though SAP competes with other big software makers, it also uses and integrates with products made by rivals including Amazon, Google, Microsoft, Nvidia, and many others.

“North America is the most competitive part of the world” for business software makers, said Lloyd Adams, head of SAP Americas, in an interview. “We are bringing more of our customers into the 4.0 center, it’s a magnet for organizations all over North America. It really helps crystallize what’s possible.”

At the center, SAP Vice President Rakesh Gandhi demonstrated how staff who remain on the factory floor wear cameras and sensors that pour data into company software systems that adjust machinery, purchasing, and pricing. “We are automating all steps,” he added. “Our big message is the importance of collecting all this information so you can seamlessly embed it with the AI and understand, real time, what is going on in your supply chain.

“The customer wants everything to run from one control center. It is important to visualize how all the technology will work. The shop floor operator can see all the analytics,” such as whether a flood in China, a strike in Britain, or a port closing in California will make it more efficient to replace one supplier with another. “They become data-driven companies.”

This story has been updated to delete a reference to Fujifilm.