Construction Holds 7% of Jobs and 38% of Heat Deaths. The Federal Rule Stalled, So an Armband Took the Shift.
Picture 3:12 on an August afternoon outside Fresno: a framer, call him Luis, feels his armband buzz against his skin. He had been setting top plates since six in the morning, drinking water when he remembered, the way every framer does on a day when the thermometer on the gang box reads 104 and the foreman is already talking about Friday's pour. His foreman's phone buzzed a second later. Then the supervisor's dashboard lit up across town. Luis sat down in the shade of the garage wall, drank a full bottle, and waited for the second vibration, the one that meant his body had cooled enough to go back up. Twenty-two minutes, and that is the whole story. A man sat down before his body made the decision for him.
Nobody passed a law to make that happen; Washington tried, then stopped trying.
How Washington shelved the heat rule
In August 2024, federal OSHA proposed the first national heat standard for outdoor and indoor workers: written heat plans for employers with more than ten people, water and shade at a heat index of 80 degrees, mandatory fifteen-minute rests every two hours at 90 degrees, acclimatization schedules for new hires, and training for every supervisor, a package that would have rewritten the morning safety huddle on every residential job in America. Labor lawyers called it the most significant worker-protection rulemaking in a generation, yet the proposal is still sitting there two years later. Public comment closed in January 2025, with the post-hearing round following in October 2025. No target date for final action appears in the latest federal regulatory agenda, and analysts at Ogletree Deakins put it plainly: the path to finalization looks unlikely anytime soon.
Meanwhile the enforcement program that was supposed to hold the line expired. OSHA's National Emphasis Program on heat, the thing that sent inspectors onto jobsites on high-heat days, lapsed on April 8, 2026, with no announced renewal. Before it died it issued sixty citations under the General Duty Clause and pulled nearly 1,400 workers out of hazardous heat, according to Ogletree's accounting, a record of enforcement that now has no successor program and no announced replacement, leaving inspectors to stretch a 1970 statutory clause over a 2026 climate. Now inspectors fall back on the General Duty Clause alone, a legal theory every defense lawyer in the country knows how to punch holes in, because a clause written for unanticipated hazards was never meant to carry the full weight of the country's fastest-growing workplace killer, year after year, with no standard behind it.
California, as usual, walked the other way. On August 14, Cal/OSHA released a revised draft that would extend high-heat procedures at 95 degrees beyond the currently covered industries, agriculture, construction, landscaping, oil and gas, to every outdoor employer in the state, which means the commissary kitchens and golf course crews get protections construction has had on paper for years, while construction itself gets a stricter acclimatization regime. It also rewrites acclimatization: close observation not just for new hires but for anyone returning after more than fourteen days away, with a choice between enhanced monitoring and a graduated exposure schedule that starts a new worker at twenty percent exposure on day one and ramps to eighty percent by day four. Seyfarth Shaw's analysis flags the message underneath: expect much greater scrutiny of how contractors bring workers into the heat.
So here is where the residential builder stands in September 2026, with the federal floor a proposal gathering dust, the California ceiling rising, and the dying continuing on schedule on actual roofs and slabs between them.
Who dies, by the numbers
Ask the building trades' own research arm and the numbers stop sounding like statistics. Construction workers are about seven percent of the American workforce and suffered thirty-eight percent of workplace heat deaths in 2020, per CPWR, the Center for Construction Research and Training. From 2011 through 2018, 104 construction workers died of heat, a count everyone in the field treats as a floor because death certificates keep writing cardiac event when the cardiac event happened on a 107-degree roof. Nearly eight in ten heat deaths land between 2 and 4 in the afternoon, June through August, which means the deadliest hazard on a residential jobsite is not a mystery chemical or a freak gust but the clock, and the clock has been testifying for a decade. WCRI, citing Bureau of Labor Statistics data, puts construction and extraction at more than thirty-two percent of all workplace heat deaths across 2011 to 2020. Specialty trade contractors, the roofers and framers and masons on residential jobs, accounted for 55.2 percent of severe heat injuries reported to federal OSHA between 2015 and 2023, CPWR reported last August.
Read that afternoon window again. Seventy-eight percent of deaths in a two-hour slot. That is not a mystery of physiology but a scheduling problem wearing a medical mask, and it is the first thing that should embarrass everyone in this conversation, the regulators who shelved the rule and the technologists selling the sensor alike.
What the armband actually does
The device is unglamorous, which is the point. A sensor straps to the forearm and records heart rate, skin temperature, and heat flux. Software combines those signals with the worker's baseline information and predicts core body temperature in real time, which is the entire trick, because everything else on the market measures the weather around the worker while this one estimates the weather inside him. When the prediction crosses the danger threshold, the band vibrates, the worker's phone alerts, and the supervisor's dashboard flags the crew member by name.
What separates this from the drawer full of dead construction wearables, the posture sensors nobody charged and the smart helmets that never survived a summer, is the validation. In a peer-reviewed 2021 study, researchers tested the algorithm against the gold standards, rectal probes and ingestible pills, across four studies and more than four thousand minutes of data. Mean absolute error of 0.25 degrees Celsius, correlation of 0.94 with true core temperature, accuracy holding from 36 to 40 degrees core temp across ambient conditions from 13 to 43 degrees. Its designers deliberately biased the algorithm to overestimate rather than underestimate, because a false alarm costs a water break and a missed alarm can cost a life. That is an engineering value judgment worth naming, because most construction tech is tuned to flatter the buyer, not protect the worker.
Big builders have run it where the liability people could watch, which tells you plenty about who this technology is really being built to reassure. Garney Construction put it on 28 workers across 10 worksites, with alerts escalating from "stop work, rest, hydrate" to emergency measures, plus an all-clear vibration when core temp returns to safe levels. Garney's safety manager framed it as prediction added to a system that already had training, hydration, and cooling stations. McCarthy Building Companies piloted it on multiple projects including utility-scale solar, where VP of National Safety Pat Devero said it was validating the company's existing heat practices, not replacing them. Even the Department of Homeland Security put $161,600 into development for its own personnel working the southern border, and the device sells as a per-worker, per-month subscription, a pricing model that tells the small residential sub everything about who the vendor imagines as the customer.
Nobody ran this comparison: the schedule is the blunt instrument
Here is the original analysis this article owes you. Line up California's August draft against the physiology paper and they contradict each other in a way nobody in Sacramento seems to have noticed.
Cal/OSHA's draft prescribes a fixed graduated schedule: twenty percent exposure on day one, ramping to eighty percent by day four, returning workers starting at fifty percent. One schedule for every body. But the validation data shows core-temperature response to heat is stubbornly individual, varying with fitness, acclimatization history, sex, and age across the tested range. A fixed calendar over-rests the already-acclimatized twenty-year-old and under-protects the fifty-eight-year-old returning after a three-week layoff, on the same crew, on the same 102-degree day, under the same compliant schedule.
A contractor running per-worker core-temp monitoring is complying with the intent of the tightening rule more faithfully than a contractor running the letter of the fixed schedule, and that gap between intent and letter is where the next decade of heat litigation will live. The state mandates the blunt instrument while the armband plays the sharp one. That inversion, the regulation as the floor and the sensor as the actual standard of care, is worth sitting with, because plaintiff's attorneys will eventually sit with it too. When the fixed schedule is the law and the precise measurement exists, "we followed the schedule" starts to sound like "we knew better and chose the average," which is an uncomfortable sentence to read aloud in a deposition and an even more uncomfortable one to explain to a jury.
On cost, here is the honest sketch. No vendor publishes per-worker pricing, so no honest break-even math is possible from public data, and you should distrust anyone who shows you some. What is public: a single OSHA serious violation draws a five-figure penalty under current inflation-adjusted maximums, before workers' compensation, before the schedule delay, before the wrongful-death deposition where someone reads your heat plan aloud. Price any subscription quote against that stack, not against zero. And price the free alternative first: moving the heaviest work out of the 2-to-4 window where nearly eight in ten deaths happen costs nothing but the courage to tell the schedule to wait.
Against the armband: the strongest case
Now the part the vendors will not put in their brochures, stated as strongly as the workers I have talked to over the years would state it.
Continuous biometric monitoring of a residential workforce that is heavily Latino, often non-union, and sometimes undocumented-adjacent in its fears is a surveillance proposition, not just a safety one. Who sees the core-temp dashboard? Picture the login list: the foreman, the safety director, the workers' comp insurer shopping for a reason to deny a claim? In 2026 those are not paranoid questions. Consent on a jobsite where the boss hands you the band with your onboarding packet is a fiction everyone agrees to maintain, signed alongside the arbitration agreement and the handbook acknowledgment, in a trailer, on day one, in a language the worker may or may not read.
Alert fatigue is real. A device that vibrates on a 100-degree roof will get ignored the way backup alarms get ignored, and an ignored sensor is worse than no sensor because the safety meeting slides will claim the crew was "monitored."
Several authors on the validation study work for the company that sells the device. That 0.25-degree figure deserves independent replication the way every vendor-funded number does. And no randomized trial shows these wearables reduce injuries or deaths compared against a disciplined water-rest-shade program, which costs nearly nothing and has decades of evidence behind it.
Hardest truth: the band measures the body with exquisite precision and does nothing about the boss, so every alert it generates is a test of whether the person with the power to stop work will stop work, and technology has never once passed that test on its own. Garney's pilot understood this, which is why the interesting part of their trial was not the sensor but the written escalation protocol behind it. Buy the protocol; the band is optional.
What this article did not prove
Several boundaries, stated plainly. Per-worker subscription pricing is not public, so the cost comparison above is a framework, not a quote; get the number from the vendor and run it against your own violation exposure. No independent randomized trial establishes that core-temp wearables reduce heat injuries or fatalities versus conventional water-rest-shade programs, and the physiological accuracy study shares authors with the manufacturer. Stalled is not dead: a future administration could revive the 2024 proposal, and its trigger thresholds could change before anything is final. California's August 2026 language is a discussion draft, and the industry-scope expansion and acclimatization provisions may be narrowed before adoption. Heat-death counts understate the problem because death certificates and OSHA logs often code heat-stroke deaths as cardiac events; the CPWR shares are floors, not ceilings. Finally, nothing here is legal advice; the Cal/OSHA draft analysis comes via law-firm reporting, and your obligations depend on your jurisdiction, your headcount, and counsel.
If you run a crew, do this
First, the free move. Audit your schedule against the 2-to-4 window. If your heaviest tasks, roofing, framing, concrete, sit inside the two hours when nearly eight in ten heat deaths occur, you have a planning failure, not an equipment shortage. Move the work or split the shift.
Second, California contractors: the August draft expands acclimatization documentation to returning workers, not just new hires. Start logging heat exposure for anyone coming back from more than two weeks off, now, before the language is final, because the contractors who survive the first inspection under the new draft will be the ones whose records already look like they expected it. Paper trails built during the draft phase are the ones that survive the inspection.
Third, if you pilot wearables, pilot them on the highest-risk crew first and demand the data policy in writing before the first band goes on an arm: who sees individual readings, how long they are retained, and whether the insurer gets access. Pair deployment with written stop-work authority for the worker wearing the band, because an alert the boss can override is a liability exhibit.
Fourth, for the ten-plus-employee shops waiting on the federal rule: write the heat plan anyway. OSHA is enforcing heat under the General Duty Clause today, stalled rule or not, and the 2024 proposal's structure, 80-degree trigger, 90-degree rest mandate, written plan, is the template inspectors already carry in their heads.
He sat down at 3:12 because something that knew his body better than he did told him to, and no regulation required it, no inspector was watching, and no federal standard will exist to require it any time soon, which is exactly why the armband matters and exactly why it is not enough. Your jobsite's standard of care is whatever you can defend after the ambulance leaves, and as of this summer, "we followed the federal rule" means "we followed a proposal." The workers knew the heat was killing them before the data did. At minimum, the industry could stop pretending the paperwork will save them.