The Manufacturing Labor Shortage Is Not Going Away

Manufacturing has a labor problem. That statement is not new, but I don’t think enough leaders are treating it as a structural issue. I compare the lack of investment in the manufacturing workforce to the deferred investment in infrastructure: the electric grid, water distribution, railways, roadways. The bill is coming due.

The Manufacturing Workforce Gap Is Already Here

At present, U.S. manufacturing employs roughly 13 million people and represents about 8% of total nonfarm employment. That is a large workforce, but it is not large enough for what is coming. Deloitte noted that manufacturing employment had surpassed pre-pandemic levels and stood close to 13 million in early 2024. More recent Bureau of Labor Statistics (BLS) data shows the sector added 15,000 jobs in March 2026, while BLS Job Openings and Labor Turnover Survey (JOLTS) data showed 439,000 open manufacturing positions in February 2026. 

That number moves month to month, but the bigger point does not move. Manufacturing is still carrying hundreds of thousands of open jobs while trying to expand capacity, reshore supply chains, adopt automation, and respond to new investment in semiconductors, clean technology, defense, and industrial infrastructure.  Additionally, consider that 2.8 million manufacturing workers will retire by 2030. That, my friends, is a lot of production capacity.

So, what is the endgame?

Are We Short Workers, or Short Skills?

The simple answer is that manufacturing needs more people. While that is true, it’s also incomplete.

Deloitte and The Manufacturing Institute estimate that the U.S. manufacturing industry may need as many as 3.8 million additional workers between 2024 and 2033. Of that need, roughly 2.8 million is expected to replace retiring workers. Another 760,000 relates to industry growth. About 230,000 is tied to legislative investment, including areas such as CHIPS-related manufacturing. The most concerning number is that up to 1.9 million of those jobs could go unfilled if manufacturers do not address both the skills gap and applicant gap.

That is not a temporary hiring inconvenience. That is a constraint on growth.

Manufacturers can buy equipment. They can build plants. They can invest in ERP, MES, robotics, AI, analytics, and planning tools. But if they cannot staff, train, and retain the people needed to run the business, those investments will underperform.

This is where the discussion often gets too vague. ​“Labor shortage” sounds like one problem. It’s really three problems.

The First Gap is Traditional Manufacturing Skills

Manufacturers still need machinists, electricians, welders, maintenance technicians, toolmakers, supervisors, assemblers, and industrial mechanics.

These are not optional roles. They keep plants running. They understand machines, tolerances, downtime, preventive maintenance, scrap, changeovers, and the practical realities that do not show up cleanly in a dashboard but do affect business performance.

Deloitte noted that production-related occupations remain the largest employment category in manufacturing and are expected to remain so. Also consider the fastest-growing production roles require higher level skills. The report also points to growth in ​“advanced manufacturing” roles, such as semiconductor-processing technicians, machinists, supervisors, welders, and electromechanical assemblers. 

That matters because many companies have spent years treating skilled trades as an aging cost center rather than a strategic asset. Now the people who know how the plant really works are retiring.

If the only response is ​“we will automate,” we are skipping a step. Automation still needs people who understand the process well enough to automate it correctly.

The Second Gap is Digital Manufacturing Skills

Modern manufacturing needs CNC programmers, robotics programmers, controls engineers, automation technicians, systems integrators, and people who understand how the digital and physical worlds connect.

This is not the same as hiring general IT resources. A plant floor is not a spreadsheet. A robot cell, CNC machine, PLC, scanner, sensor, ERP transaction, and production schedule all have to work together in a messy operating environment. The missing link is the person that understands what each handoff should look like and what it means.

Manufacturers are moving toward smart factories and Industry 4.0, but the labor market is not producing enough people who can bridge that gap. Deloitte’s research states that role requirements are changing as manufacturers move toward smart factory operating models, and that demand is increasing for workers with technical manufacturing, digital, and soft skills.  Soft skills like change management are essential, because this shift represents a massive change and with change, comes conflict.

The issue is not just whether a person can program a machine. It is whether they understand what the program is supposed to accomplish in the context of throughput, quality, costing, maintenance, scheduling, and customer demand.

That is a harder profile to find and develop, especially if a company chases the bright shiny penny of AI to fill the entry level roles.

The Third Gap is Information Skills

This may be the least obvious shortage, but it may become one of the most important.

Manufacturers need analysts, data scientists, logisticians, planners, and finance resources who understand operations. They need people who can interpret the information created by the plant, not just report it.

That means understanding inventory behavior, routing accuracy, labor standards, machine utilization, supplier constraints, costing assumptions, lead times, and forecast error. It also means knowing when the data is wrong.

Deloitte found that software developers, computer and information systems managers, and computer and information analysts already represented a meaningful manufacturing employment base in 2022 and could grow nearly 13% by 2032. Statisticians and data scientists remain a small share of manufacturing employment, but those roles could grow close to 30% by 2032. 

Those numbers should make manufacturing leaders pause. 

The plant of the future will not only be run by people who can make parts. It will also be run by people who can interpret signals.

Can Robotics Replace the Missing Workforce?

Robotics will help. It has to.

For traditional manufacturing tasks, robotics can reduce repetitive work, improve consistency, and protect output when labor is scarce. For dangerous, dirty, or ergonomically difficult work, robotics is a strong solution.

But robotics does not eliminate the workforce problem. It changes it.

A robot still needs to be justified, specified, installed, programmed, maintained, scheduled, measured, and improved. If the plant lacks the people to support that lifecycle, automation becomes another underperforming asset.

This is where apprenticeship programs matter. Manufacturers need structured ways to bring in new workers and build skills over time. Not a loose onboarding checklist. Not tribal knowledge passed down only when someone has time. A real training path.

The problem is that apprenticeship programs require patience. Many companies want the benefit of skilled labor without making the long-term investment required to create it.

That math does not work.

Can AI Replace the Information Gap?

AI can help here too, but we need to be careful.

AI can summarize data, identify patterns, draft explanations, generate first-pass analysis, and make information easier to access. In planning, reporting, procurement, maintenance, and FP&A, that can be valuable.

But I would not confuse faster output with deeper understanding.

A junior analyst using AI may produce work faster. A plant manager using AI may get an answer faster. A scheduler may get a recommendation faster. That is useful. But who validates the answer?

If the people using AI do not understand the underlying process, they may accept a clean-looking answer that is operationally wrong. Manufacturing does not forgive that for long. Bad assumptions eventually show up as missed shipments, excess inventory, bad costing, poor labor planning, or margin erosion.  If the junior roles that have been producing the data and developing the understanding of the cause and effect in the data are replaced with AI agents, when more senior team members leave, that knowledge leaves with them.

AI is best used as an accelerator for people who are learning and a force multiplier for people who already know the business. It is risky when used as a substitute for building judgment.

What is the Risk of Automating Too Quickly?

The first risk is a broken professional growth path.

If manufacturers automate entry-level work without creating new ways for people to learn the business, they may solve today’s labor issue by creating tomorrow’s leadership issue. The person who becomes a strong maintenance manager, operations leader, plant controller, planner, or manufacturing engineer usually starts by learning the details. If those details are hidden behind systems and algorithms, the next generation may not get the experience needed to lead.

The second risk is dependency.

Automation, robotics, AI, and advanced systems require power, connectivity, vendors, spare parts, cybersecurity, and technical support. That dependency is not a reason to avoid automation. It is a reason to be sober about it.

The more automated the plant becomes, the more fragile it can become if resilience is not designed in. A plant that cannot run when the network is down, when the power is unstable, or when one specialized technician is unavailable has not eliminated risk. It has changed the form of risk.

So What Do Manufacturers Need to Do?

First, stop treating labor as only an HR issue. This is an operating strategy issue.

Second, separate the gaps. Traditional skills, digital skills, and information skills require different recruiting, training, compensation, and retention plans.

Third, build apprenticeship and mentorship paths that are tied to actual plant capability. Don’t just train people to fill roles. Train them to understand the business.

Fourth, use robotics and AI aggressively, but not blindly. The goal should be to improve productivity while increasing organizational knowledge, not hollowing it out.

Finally, measure workforce capability the same way manufacturers measure production capability. If a plant knows its OEE, scrap rate, labor efficiency, and backlog, it should also know where knowledge is concentrated, where retirements create exposure, and where digital skill is too thin to support the automation roadmap.

The labor shortage is not just about finding workers. It is about deciding what kind of manufacturing company you want to be ten years from now.

If the answer is a company with more automation, more data, more AI, and more advanced equipment, then the workforce strategy must become more advanced too.

Otherwise, the machines may be ready, but the business will not be.