How Is AI Used in Manufacturing? A Practical Guide to AI Tools for Manufacturing

Not long ago, artificial intelligence was mostly a boardroom talking point for manufacturers, more theory than practice. That has changed fast. Over the last 12 to 18 months, manufacturers have moved from cautiously discussing AI to actively deploying it on the shop floor, in the back office and across the sales team.

Rather listen than read? Watch the full on-demand webinar and hear directly from the experts.

Three forces are driving the shift.

Doing nothing is no longer a neutral choice. It runs the risk of falling behind competitors in efficiency and losing employees to organizations that offer more modern, digitally enabled tools to do their jobs.

How is AI Used in Manufacturing and Where is it Delivering Value?

The strongest results tend to cluster around three areas.

From tariffs to inventory management, Withum provides the guidance and operational expertise to help manufacturers grow.

Click to learn more about Withum’s Industrial and Consumer Products services.

What’s the Biggest Hurdle to AI Adoption in Manufacturing?

Ask manufacturers where they get stuck and the answer usually isn’t the technology, it’s knowing where to start. Identifying the right use case remains the most common challenge, ahead of data quality and integration, with security and governance concerns further down the list. In other words, the barriers are organizational.

Where Should Manufacturers Look for Their First AI Use Case?

Nearly every organization has a “Bernice.”

“Look for Bernice, that poor soul in your organization who’s been there for 13 years,” says Andrea Mondello, who leads Withum’s AI Services practice. “She spends two weeks out of every month taking one spreadsheet from one system, doing stuff to it and handing it off to somebody else, who puts it into another system. And most of the time, the C-suite doesn’t even know Bernice exists.”

That gap between what leadership assumes is happening and the manual work that actually holds a process together is often the best place to start. It’s low-risk, the payoff is obvious once it’s fixed and finding it tends to reveal exactly how ready an organization’s data really is before a bigger AI investment gets underway.

How Should Manufacturers Evaluate an AI Investment?

Every AI initiative eventually becomes a financial conversation, and it should be treated that way from the start. Rather than asking where AI could be used, leaders should start with the business problems they are already struggling with and work backward. From there, a structured business case should outline expected savings, revenue impact and effect on headcount, applying the same discipline used for a new piece of equipment or a building purchase.

“Treat it just like you would treat a machine or any other major investment,” says Walter Merkas, who co-leads Withum’s Digital Workplace Solutions and Management Consulting team. “Ask your teams to quantify the plan and the return, because the AI cost itself is often limited. It’s the internal staff time, the licensing and the data cleanup that get underestimated.”

Costs go beyond software licensing. Internal staff time and data cleanup are often underestimated, and any ROI analysis should account for both. The payoff can be fast: workflow automation, in particular, tends to deliver a quick payback because it removes redundant manual touches on the same document or task.

What Mistakes Are Manufacturers Making with AI?

Three missteps come up repeatedly.