Supply Chain Expert Explains Why AI Pilots Stall at the Second Plant in Industry Today

Rob Williams, supply chain lead at Centric Consulting, wrote an Industry Today byline on why manufacturing AI pilots that succeed at one plant often stall when they expand to the next.

In “Why Successful AI Pilots Struggle to Scale,” Williams looks at what happens after a manufacturing AI pilot proves itself at one plant and the company tries to replicate that success elsewhere.

That first success can be misleading. Williams explains that first-site pilots usually run on a single plant’s data, which is often cleaned by hand. That lets them skip the enterprise data infrastructure a wider rollout needs. “The reason it works has less to do with technology than scope,” he wrote.

Those shortcuts break down at the next plant, where the equipment, data labels, and systems are different, so the pilot can’t be copied over as is. “What looked like a repeatable win at the first site becomes a custom project at the next,” said Williams.

That tension between enterprise consistency and plant-level flexibility also appears when manufacturers map value streams across multiple facilities. “Scaling works when the data model comes first and the pilot is treated as one use case built on top of it,” he wrote.

To address those challenges, Williams provides advice on building a foundation for scaling, planning each site’s rollout, and balancing enterprise standards with plant-level flexibility.

Read the full article in Industry Today for Williams’ guidance on scaling successful AI pilots across manufacturing plants.