10

Loading video...

Michael Holmes

By Michael Holmes

0

Share

Newsweek is a Trust Project member

See more of our trusted coverage when you search.Prefer Newsweek on Googleto see more of our trusted coverage when you search.


The next test for American manufacturing is whether artificial intelligence can help people make better decisions under real production pressures. A factory needs reliable output, consistent quality, and workers who know when something is wrong. I believe AI investment should strengthen that judgment, with the people doing the work helping decide how the technology is used.

The choice is becoming practical. Deloitte’s 2025 Smart Manufacturing Survey, covering 600 executives at large companies with U.S. headquarters or operations, found that 29 percent reported using AI or machine learning at the facility or network level.

As adoption develops, manufacturers have an opportunity to define success beyond the number of tasks automated or positions removed.

Read More on Leadership Insights

Put Workers First

Starting with a headcount target risks overlooking the work that makes production dependable. An operator may need a maintenance record, a quality requirement, or an explanation of why a procedure changed.

Making that information easier to find could improve a decision. Removing the person who understands its significance could weaken it.

That is why shop-floor employees should help shape an AI project before purchasing decisions are finalized. Operators, maintenance managers and team leads should identify where information is missing, what slows a shift and which decisions require experienced judgment.

File image of an industrial factory.

...

Their involvement gives a manufacturer a clearer problem to solve and a practical way to judge whether the investment helps.

My approach is simple: AI multiplied by human decision. The technology should retrieve and organize relevant information; a qualified person should interpret it and remain responsible for consequential choices. The purpose is to expand what a worker can accomplish. Faster access to information matters when it supports a sound decision, not merely when it produces a quick answer.

Consider a hypothetical operator investigating a recurring equipment fault. An AI tool could help bring together approved instructions and earlier maintenance notes.

The operator would still need to check whether those records apply to the machine’s present condition and involve maintenance staff when necessary. A useful system should make its sources visible and allow uncertainty to be raised without pressure to accept its answer.

Preserve Experience

Preserving experience matters just as much as retrieving documents. McKinsey’s July 2025 analysis of the manufacturing workforce identified knowledge lost as experienced employees retire as a productivity concern. A searchable manual alone cannot preserve everything an experienced worker knows.

Manufacturers should give senior employees time to explain exceptions, document past decisions, and review the material newer colleagues will use.

AI could help organize those records into accessible training resources, but experienced staff should check their accuracy and relevance. New employees also need supervised practice. The aim should be to help them understand why a process works, how it evolved, and when to ask for help.

Access to information should support the development of competence, rather than be mistaken for competence itself.

Measure More Than Efficiency

Efficiency also requires a workforce plan. In 2023, Ingka Group reported reskilling 8,500 call-center employees in areas including remote interior design and digital sales while its chatbot handled simpler inquiries. This was a retail example, not evidence of factory results.

It nevertheless illustrates a choice manufacturers should consider: preparing employees for different responsibilities as technology changes their work.

Having worked in textiles and furniture manufacturing for almost 30 years, I have seen production move offshore. That experience shapes my concern about competitiveness and what happens to workers when businesses change. Manufacturers cannot assume that protecting existing routines will protect jobs.

Equally, they should not assume that reducing payroll is the only worthwhile return on technology. More reliable production and stronger skills deserve a place in that calculation.

Before expanding an AI project, operations leaders should test it with the employees who will use it. Measure whether it helps resolve a defined production problem, protects quality, and supports sound decisions. Assign responsibility for checking its information, preserve workers’ authority to challenge it, and fund training for any changed roles. If efficiency frees up time, identify useful work people can learn to do.

The standard I would set is straightforward: demonstrate how the investment strengthens both production and the people responsible for it. American manufacturing needs technology that leaves its workforce more capable of making the next difficult decision.

Michael Holmes is the Chief Consulting Officer and owner of HMA Consulting Group and HMA Tech LLC. The views expressed in this article are his own.

View 0 Comments

Request Reprint & Licensing

Submit Correction

View Editorial & AI Guidelines