On a recent episode of Advanced Manufacturing Now, study authors Steve Shepley of Deloitte and Victoria Bloom of the Manufacturing Institute laid out the case that AI — deployed correctly — can stretch the existing workforce further and lower the barriers for new workers to become productive faster.
The study breaks technicians into three categories: advanced manufacturing technicians who support design, testing, calibration, and optimization of equipment and systems; maintenance and repair technicians who keep fielded equipment running through preventive maintenance and troubleshooting; and quality and laboratory technicians focused on testing, analysis, and validation. All three are in acute demand. All three draw from talent pools that other industries — telecommunications, heavy equipment, forensics, agricultural research — are also fighting to recruit from.
That competition for talent is part of the study’s thesis. Millions of workers in adjacent industries already carry transferable skills. The gap between what a fiber optics technician knows and what an advanced manufacturing technician needs to know is narrower than most hiring processes acknowledge. AI, the authors argue, can compress the reskilling timeline and make cross-industry mobility more practical.
The study’s most compelling evidence comes from a real-world maintenance example. A manufacturer studied its top-performing field technicians and found that their approach to troubleshooting varied dramatically from the company’s standard operating procedures. The difference wasn’t talent — it was experience accumulated over five to ten years of servicing equipment across different environments and conditions.
The company trained a large language model on that accumulated knowledge — historical service records, usage data, fault codes, environmental context — and deployed it on handheld devices. Field technicians could upload photos and videos of the issue, answer a few contextual questions, and receive a customized corrective action procedure generated on the fly. The result was a 30 to 40 percent reduction in escalations to senior technicians, achieved in initial use cases alone.
Rather than needing to learn and wait ten years to build that experience level, they’re actually armed with it today. We found that the delivery of AI on handheld devices reduced escalations to the most senior technicians by about 30 to 40 percent.
The point isn’t that AI replaces the experienced technician. It’s that AI captures what the experienced technician knows and distributes it to everyone else. The senior tech still validates. The model still learns from feedback. But a newer hire troubleshooting a piece of equipment in a corrosive coastal environment no longer starts from the same generic SOP as someone servicing identical equipment in a climate-controlled facility.
The study surfaces a familiar disconnect: 85 percent of survey respondents said building a more agile, adaptive workforce is very important. Only 7 percent said they’re making meaningful progress. Shepley attributes the gap partly to the pace of model evolution — companies invest in building a solution around one AI model, and three months later a newer model is available. Experimentation is showing promise at many manufacturers, but few have reached scale with a fully refined implementation.
The study outlines five actions for manufacturers looking to close that gap: develop a prioritized AI implementation roadmap, redesign technician workflows around human-AI collaboration rather than layering AI onto existing systems, define AI-enabled skill requirements and rebuild training pathways accordingly, embed continuous learning into daily work instead of treating it as a separate change-management exercise, and strengthen partnerships with educational institutions to modernize technician workforce development.
The technology is outrunning the workforce’s readiness to use it. The Manufacturing Institute is trying to close that gap through a new AI skills initiative funded by a $10 million grant from Google. The program offers free online courses built around real manufacturing use cases — not generic AI education — covering fundamentals, responsible AI practices, and practical daily applications. The initiative is also being integrated into the Institute’s FAME apprenticeship program, which partners with more than 450 employers nationwide, so new technicians build AI fluency from the start of their careers rather than bolting it on later.
Deloitte is pushing even further upstream with its Believers program, a workforce development initiative that puts hands-on STEM resources — smart rover kits, customized mechanical equipment — into high school classrooms. The goal is career exposure early enough to reduce dependency on four-year degrees as the only pathway into technical roles.
The question’s not whether manufacturers will use AI, but whether their workforce is prepared to leverage it. Executives and frontline leaders are going to have to play a role in making sure their workforce has the tools to utilize it and understand the benefits for them.
Shepley frames AI alongside the railroad, the assembly line, the computer, and the cell phone — technologies that felt transformative in their early innings and became infrastructure within a generation. Manufacturing is in the first few innings of that cycle. The companies that invest in workforce readiness now won’t just have a head start when the technology matures. They’ll have technicians who know how to work with it — which, given the shortage numbers, may be the more durable competitive advantage.
The 2.3 million technician openings projected through 2030 aren’t going to be filled by recruiting alone. They’ll be filled by making the workers who are available more capable, faster. That’s not a vision statement. It’s a 30 to 40 percent reduction in escalations, measured in the field, today.
