The Industrial Workforce Crisis Nobody Wants to Price Correctly

I just finished binge watching season two of Landman on Paramount+. Couldn’t stop. Partly Taylor Sheridan knows how to end every episode on a knife’s edge, but mostly because I lived season one. Minus the cartel.

The long hours. The roads that don’t have names, just coordinates. The 2 AM calls that don’t care about your schedule. The constant, grinding, unglamorous problems that stack up without a break. And the rattlesnakes. In the show, the lawyer stood there frozen. In real life, I ran my ass to the truck while the pumper grabbed his shovel.

But what the show gets right, maybe without meaning to, is the question underneath all the drama. Not the lease negotiations or the pipeline politics. The simpler, harder question that every character eventually faces: why would anyone choose this?

Because increasingly, they don’t.

The Math

A roughneck on an onshore rig in the Permian averages about $48,000 a year. A retail store manager at a big box chain averages $68,000. One of them works 12 to 14 hour shifts on a steel platform in 110 degree heat, two weeks on, one week off, in an industry with a fatality rate more than four times the national average. The other argues about return policies in air conditioning and goes home at six.

This is not an abstraction. This is the actual decision facing a 22 year old in Midland or Odessa right now. And the decision is getting easier every year, in the wrong direction.

Oil and gas extraction recorded 32 fatalities in 2024. Construction killed 1,034 workers, roughly one in five of all U.S. workplace deaths. Waste collection is the fifth deadliest industry in America. Between 2015 and 2022, oil and gas alone saw 2,101 severe injuries: amputations, eye loss, hospitalizations. These are not rounding errors. These are the terms of employment.

Seventy six percent of Gen Z says they prioritize work life balance over pay. A third say they’d leave the trades entirely if conditions don’t improve. You can call that soft. Or you can call it rational. Either way, the pipeline is thinning.

The Squeeze

The workforce that built and maintained America’s industrial infrastructure is aging out at a pace that has no precedent. 4.1 million Americans are turning 65 every year from 2024 through 2027. A quarter of the manufacturing workforce is over 55. The average age of an oil and gas worker is 56. In water utilities, a third of all operators can retire within the decade, and only 10 percent of the workforce is under 24.

Those numbers would be alarming on their own, but the boom and bust cycle in energy made it worse. In 2014, more than 600,000 people worked in oil and gas extraction. Today it’s roughly 380,000, producing 47 percent more oil with 37 percent fewer people. When the price crashed in 2015, workers scattered. When COVID hit in 2020, 107,000 more jobs disappeared in five months, the fastest layoff in the industry’s history. Most of those people found work in logistics, retail, warehousing. Stable hours, predictable income, nobody gets crushed by pipe. They didn’t come back.

The bust doesn’t just shrink the workforce. It poisons the recruiting pitch for a generation. Why train for a career that might lay you off every five to seven years? Petroleum engineering enrollment dropped 15 to 21 percent between 2015 and 2019. It hasn’t recovered.

The bust poisoned something else too: trust. When the boom was running hot, a wave of startups flooded the oil patch. A lot of them were chasing the Hollywood version of entrepreneurship, venture funded, big promises, slick decks. When the bust came, most of them vanished. They never delivered. And the operators who’d taken a chance on them got burned. That made the next conversation harder for everyone, including us. The reality is that you have to grind with these operators to gain their trust. You have to prove it works on their site, in their conditions, and pivot with them when things change. There are no shortcuts. The startups that treated the oil field like a quick market opportunity left a credibility hole that the rest of us are still climbing out of.

What Walks Out the Door

The headcount problem is visible. The knowledge problem is invisible, and it’s worse.

I once watched a pumper walk up to a wellhead, press the handle of a wrench against the casing, and listen. After about ten seconds he pulled out his phone, called the engineer, and told him exactly what was failing downhole. The engineer pushed back. The pumper argued. The pumper was right. He’d been listening to that well for fifteen years. He didn’t need a sensor. He was the sensor.

An estimated 70 percent of critical operational knowledge in industrial settings is exactly that: tribal. Never written down. Never formalized. It lives in the heads of people who’ve been doing the job for 20 or 30 years. They know which vibration pattern on a compressor means a bearing is going. They know which color shift in a waste stream means contamination is spiking. They know which valve to check first when a reading looks wrong, not because a manual told them, but because they’ve seen it fail three times in 1997, 2004, and 2011.

When they retire, that knowledge doesn’t transfer. It vanishes. The U.S. government learned this the hard way when it discovered it could no longer manufacture a critical component for nuclear warheads. The engineers who knew how had all retired. It took $69 million and five years to relearn what a handful of people once carried in their heads.

Eighty three percent of water utilities acknowledge that tribal knowledge is lost when a worker leaves. Research estimates that knowledge loss costs organizations $47 million per year in errors, extended training, and duplicated problem solving. Multiply that across every water plant, every refinery, every waste facility, and every pipeline operation in the country and the number stops being a statistic and starts being a structural failure. This is happening now, at scale, across every sector that touches physical infrastructure.

Careers Aren’t Enough. We Need a Calling.

The industry’s response so far has been to recruit harder. More job postings. Higher signing bonuses. Marketing campaigns about the dignity of skilled trades. And it’s not nothing. Gen Z enrollment in trade programs increased 23 percent in 2024. Vocational community college enrollment is up 16 percent. Siemens committed to training 200,000 electricians and manufacturing experts by 2030. Those are real numbers moving in the right direction.

But they’re moving against a gap measured in millions. 2.1 million manufacturing jobs projected unfilled by 2030. 349,000 construction workers needed this year alone. 300,000 water utility positions in the next decade. The Department of Education says there are 68 percent more job openings in infrastructure fields than people training to fill them.

Recruiting campaigns won’t close that. Not at current pay. Not at current risk levels. Not in an economy where a kid with a laptop and a wifi connection can make more money doing social media management than a lineman who climbs utility poles in ice storms for a living.

What might work is something closer to a calling. In 1933, FDR created the Civilian Conservation Corps and put three million young Americans to work building the country’s infrastructure. They constructed 126,000 miles of roads, 45,000 bridges, and planted 2.3 billion trees. It worked not because the pay was good. It worked because the mission was clear and the purpose was larger than the paycheck. The Peace Corps borrowed from the same playbook. So did the GI Bill.

The American Climate Corps took a step in this direction with 9,000 participants in its first year. But 9,000 people against a gap of millions is a gesture, not a solution. If we’re serious about rebuilding aging water systems, modernizing the grid, maintaining the energy infrastructure that keeps the economy running, we need to talk about national service at a scale that matches the problem. Not as a fallback for people who couldn’t get a desk job. As a genuine call to build something that lasts.

Changing the Equation

There’s a parallel path that doesn’t require waiting for Washington. AI and continuous monitoring can’t solve the headcount problem. But they can change the knowledge problem.

If a camera with edge inference can detect what a 30 year veteran detected by instinct, the replacement worker doesn’t need two decades of pattern recognition to be effective. They need training on the system, not on the lore. Predictive maintenance using edge AI has already cut unplanned downtime by 40 percent in leading plants. Augmented worker platforms are giving technicians with two years of experience access to decisions that used to require twenty.

This is the work we do at Andium. We build edge AI devices that monitor industrial sites continuously, capturing the kind of visual and thermal intelligence that used to exist only in the heads of people who aren’t going to be around much longer. The goal isn’t to replace the veteran. It’s to make sure what the veteran knew doesn’t disappear when they do.

But technology alone isn’t sufficient. Ninety five percent of U.S. industrial businesses plan to introduce new automation within three years. That’s a supply response. The demand side, the question of who actually shows up to do the physical work that automation can’t reach, still requires people. And attracting those people requires honesty about what the work pays, what it costs, and why it matters.