I spend much of my time these days in rooms with plant heads, automation leads, system engineers and almost every conversation about “AI-ready manufacturing” starts with machines and ends with people. We talk about sensors, digital twins, predictive models, robotics on the shop floor and then someone asks, “But who’s going to run all this?”
For manufacturing organizations, workforce transformation is becoming a fundamental part of the infrastructure required to support AI-ready operations. It isn’t a secondary HR function or a training line item – it is as critical to performance as hardware on the shopfloor or the datapipes running through it.
The math doesn’t say what people think it says
Deloitte’s 2026 Manufacturing Industry Outlook estimates that more than 81% of manufacturing task hours will continue to be human-driven, even as AI adoption is expected to roughly double, from 9% to 22%, over the next couple of years.
We’re heading into one of the fastest periods of automation adoption manufacturing has seen, and the overwhelming majority of the work is still going to run through human hands and human judgment. That’s not a contradiction. It’s the actual nature of the transition. AI isn’t replacing human oversight – it is shifting what operators need to master.
Workforce strategies built on the assumption that automation shrinks the need for skilled people are fundamentally flawed. What shrinks is the room for narrow, single-discipline expertise. What grows is the need for talent that can can move fluidly across disciplines.
Multidisciplinary talent is the real differentiator
Here’s what I mean by that. A connected factory today needs people who understand mechanical and electrical engineering but also fiber and network connectivity, because a smart factory runs on the same reliability principles as any other critical network. It needs people who are fluent in AI and automation to work alongside intelligent systems.
And it needs people who can read data, not as analysts, but as operators, quality engineers and maintenance leads who treat data literacy as a core job skill, the same way you’d expect them to read engineering schematics.
Manufacturers who get this right are the ones who build teams capable of maximizing hardware to its potential, troubleshoot it, question it when the model gets something wrong, and improve it over time. That capability doesn’t appear on a spec sheet, but it shows up in uptime, in quality, improved productivity and in how fast an organization can adapt when the technology itself changes, which in the AI era – it will, constantly.
That’s the competitive edge. Not “who has the most automation,” but “whose people can do the most with it.”
Learning has to become continuous
For most of my career, technical training followed a periodic cadence – onboarding, a milestone certification, or an occasional refresher. That model no longer fits the pace of modern manufacturing anymore. The technologies underpinning intelligent manufacturing are moving fast enough that a skill set from eighteen months ago can already be dated.
We’ve had to rethink learning as an ongoing operational rhythm rather than a discrete event. That means shorter, more frequent upskilling cycles tied to what’s actually changing on the shopfloor, not a generic annual curriculum.
It means giving engineers exposure to AI tools and data fundamentals even if that’s not their “core” job, and giving digital and IT teams enough grounding in production realities – uptime pressure, safety constraints, quality tolerances.
This is a permanent state, not a phase we’re passing through on the way to some steady future. Continuous learning is the new steady state.
Breaking silos is an operational necessity
I’ll be candid: “break down silos” is one of those phrases that’s been said so often it’s started to sound like filler. But in AI-ready manufacturing, it’s not a nice-to-have culture goal. It’s a functional requirement.
An AI-enabled production line doesn’t recognize the line between “manufacturing”, “IT” or “digital transformation.” A single sensor feeding a predictive maintenance model impacts production schedules, data engineering, and IT infrastructure simultaneously. If these teams operate in isolated reporting lines with separate priorities and separate vocabularies, the technology will underperform no matter how good it is on paper.
The organizations building genuinely resilient, connected operations are the ones actively engineering collaboration between these functions like shared goals, shared data, and shared accountability for outcomes.
Building the next decade
None of this is a call to slow down on technology adoption. It’s a call to treat workforce capability with the same seriousness we bring to any other major infrastructure investment. You wouldn’t deploy a connected factory on unreliable power or unreliable connectivity. You shouldn’t deploy one on an unprepared workforce either.
The manufacturers who’ll lead the next decade won’t be the ones who automated the fastest. They’ll be the ones who built people capable of making AI actually work on the ground; multidisciplinary, continuously learning, and organized around the problem rather than around the org chart.
That’s the infrastructure I think about most. And it’s the one I’d encourage every manufacturing leader to start investing in now, before the technology outpaces the people meant to run it.
We’ve featured the best AI website builder.
This article was produced as part ofTechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.
The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here:https://www.techradar.com/pro/perspectives-how-to-submit
