In the rush to adjust sourcing strategies in line with current trends in international trade, the answer might be to think small.

U.S. manufacturers have spent the last few decades consolidating production at gigantic offshore plants, especially in China. But for many, rising tensions between trading partners, coupled with higher labor and logistics expense, have made that strategy untenable.

The solution for some lies in bringing production closer to American consumer markets, if not all the way home. It might mean setting up plants in Mexico or elsewhere in Latin America, aa means of cutting transit time and responding more quickly to changing patterns of demand. It could even entail some degree of domestic production.

What it doesn’t necessarily mean is simply transporting those Chinese mega-factories to another hemisphere. Such a model, based on prioritizing unit economies and production at scale, doesn’t prepare manufacturers and retailers for the waves of disruption that are washing over global supply chains today.

Enter the concept of the “microfactory.” As the name suggests, it’s a much smaller production facility that promotes better sensitivity to the characteristics of local markets.

If a typical automotive plant is turning out a million or more cars in a year, a corresponding microfactory might be producing 10,000 to 20,000 units, notes Anupam Singhal, president of manufacturing with Tata Consultancy Services. Not only does such a setup give producers “a better feeling about the market,” he says, it could also protect them from punishing tariffs.

The modern-day microfactory would not, however, be possible without the advent of modern-day technology — in particular, artificial intelligence. Singhal says AI is instrumental today in helping buyers to locate suppliers that can provide maximum agility. They can make selections based on an optimal mix of price, quality and geographic location. And they can identify alternative sources in the event of a disruption in the primary supply line, whatever the cause.

AI also can aid in speeding up the whole product-development process, from ideation to design, testing and full-scale production, Singhal says. Agentic AI, in particular, can be deployed to address problems based on specific customer use cases. “It can bring quality to a high level at a much lower cost,” he says.

All of that theoretically lends itself to the creation of smaller, more nimble facilities. But to make sure that they stay in tune with inevitable shifts in the market, as well as any number of unanticipated supply chain disruptions, manufacturers have yet another technology tool at hand: the digital twin.

Variations on the digital twin — a virtual representation of an entire physical operation — have been around for well over a decade. When combined with the latest advances in AI, however, they can be deployed at an unprecedented scale and degree of accuracy. Manufacturers are able to envision multiple “what-if” scenarios before making decisions on production locations, levels and layout. They can also envision the fallout from an outright shutdown of operations, and have alternative strategies at the ready.

Automation of key decision-making dramatically lessens the need for intervention by humans carrying out tasks manually. With the help of agentic AI, “perhaps human intervention could be limited to 5% to 10%,” Singhal says. Considering the cost of American labor, that could make the notion of localized microfactories even more feasible.

Further contributing to an efficient microfactory is 3D printing, another technology that has been around for a good while, but could find a new level of acceptance as producers seek alternatives to traditional manufacturing environments.

Localized microfactories won’t come about with the wave of a metaphorical magic wand, however, even with the help of cutting-edge technology such as AI. There will still be the need for multiple tiers of suppliers, many of which would remain offshore, Singhal acknowledges. And the entire concept remains in its infancy.

“The conversations are becoming real,” he says. “You want to think about how to sustain manufacturing without being held to the geopolitical situation.”

What’s certain is that a new approach to sourcing and manufacturing is required to succeed in a world where supply chain disruption is the norm. And AI promises to play a key role in that effort over the long term. Generative AI, for one, has only been on the market for two or three years.

“I think there’s something new,” Singhal says. “There’s much more of an appetite to adopt than in the past.”

NEXT: More on AI and 3D printing in manufacturing.