SAFE verdict contested
The core of this job is walking the rail yard, climbing into aircraft wheel wells, crawling under trailers, and physically verifying brakes, welds, seals, hazmat placards and cargo securement — sensing and access problems robotics still handles badly. AI eats the paperwork half: violation write-ups, DOT/FAA form population, defect-trend analysis, and pre-screening which carriers to audit. What holds is the legal authority behind an inspector's signature — an out-of-service order or airworthiness sign-off requires a credentialed human who owns the call.
This fall is concentrated in 2020 and has not recovered since.
Median pay $75,820 → $92,100 -2.8% in real terms
This line is counted by the Bureau of Labor Statistics — the one figure on this page that isn't a judgement of ours. Headcount moves on demand, offshoring, demographics and the business cycle, and automation is one term among several, often not the loudest.
So a falling line is not evidence that AI did it, and a rising one is not evidence that it won't. Both happen in this register: some occupations resist automation and shrink anyway, others are highly automatable and keep growing. The marked year is 2020.
BLS projection, 2024–2034
+1.7% 25,700 → 26,100 on the projections basis
Hard to automate, and growing
The work resists current AI and the BLS projects +1.7% more of these jobs by 2034. Note that safe does not mean well paid — several of the fastest-growing resistant occupations are among the lowest paid on the register.
Different clocks. The score is what current AI could do to this work today. The projection is how many of these jobs will exist in 2034. Everything between the two — how fast employers actually adopt, whether demand grows in the meantime — is why they can point opposite ways without either being wrong.
~2,500 openings a year on average, including replacing people who leave.
CarmanInspectorAdmeasurerCar ExaminerBus InspectorCar InspectorJet InspectorPit InspectorShip SurveyorWay InspectorAuto InspectorCargo SurveyorSafety OfficerTank InspectorCargo InspectorMarine SurveyorNaval InspectorSafety EngineerSmog TechnicianState InspectorTrack InspectorTrain InspectorWatch InspectorWheel Inspector
Holding it up: embodiment . Weakest point: trust premium .
Tasks largely resist digitisation A CVSA Level I inspection requires physically pulling wheels off to measure brake lining thickness, tugging on air lines, and eyeballing frame cracks under a trailer — the 15 rather than 18 reflects that the screening layer (PRISM/ISS carrier risk scoring, ELD data pulls, defect-history triage) is already algorithmic and deciding who you inspect before you ever walk up to the truck.
Hands-on in uncontrolled environments You work in live rail yards, on aircraft ramps, at roadside scale houses in weather, climbing ladders onto tank cars and reaching into landing gear bays — uncontrolled sites with moving equipment, which is 17 and not 20 only because a meaningful share of the shift is spent in the truck or the office closing out reports.
Licensed human required and personally liable An FAA A&P/IA airworthiness release or a DOT out-of-service order carries a named credentialed signature and personal exposure under 49 CFR and 14 CFR — a 14 rather than 18 because many state and local vehicle inspectors work under agency certification and delegated authority rather than an individually revocable professional licence.
Meaningful discretion Deciding whether a hairline weld crack, a marginal brake stroke measurement, or an improperly blocked hazmat load justifies pulling equipment out of service is a judgment call with real economic and safety consequences and no lookup table that resolves it — held at 13 because inspection criteria (CVSA out-of-service criteria, FAA ADs) constrain the range of defensible answers more tightly than in most discretionary roles.
The verdict above describes this occupation as a whole. Almost nobody does the typical version of a job — tick what's actually in your week and see how your own mix sits.
Your task mix speaks to task resistance (15/20 here) — how much of the day's work current AI already does. That is the dimension the boxes above are about.
It cannot move the other three. Liability shield (14/20) is whether the law requires a licensed human to sign. Trust premium (8/20) is whether buyers specifically pay for a person. Judgment and accountability (13/20) is whether the role exists to own consequential calls. Those are facts about the occupation's standing, not about which tasks are in your week — a paralegal who does only trial exhibits still holds no licence. Together they are 35 of this occupation's 67 points (52%).
Embodiment (17/20) is also a property of the work rather than the worker, but we don't tag individual tasks as physical or not, so the picker can't tell you anything about it. That's a limit of this tool, not a claim.
Did we get the list right? Tell us what's missing — the tasks are written from the outside, and you're reading this from the inside.
The moves above are yours to make. This is the other half: what would have to change in the world for the occupation itself to score higher. None of it is in any one person's gift, but it is where the floor actually comes from. Scores here are not a one-way ratchet. Only two of the five dimensions — task resistance and embodiment — track what machines can do. The other three track law, what buyers will pay for, and who is answerable, and those move in both directions, often in response to the same pressure AI creates. If every lever below landed, this occupation would score around 78/100, still SAFE.
Task-mix shift: this job genuinely has two tiers. If form population, defect-trend analytics and carrier risk pre-screening (already being built into FMCSA's CSA/SMS and airline MRO analytics) are fully automated, the residual role concentrates on adjudicating ambiguous physical findings — borderline weld cracks, brake stroke measurement disputes, contested securement calls, and carrier appeals — which resists automation more than the average of today's task set.
Expansion of statutory physical-inspection frequency into settings robots handle worst: enactment of the periodic-inspection and two-person crew provisions from the Railway Safety Act of 2023 (S.576) style bills, or state laws like the wayside-detector and walking-inspection mandates passed in Ohio and Nebraska, which specify a human walking the train rather than accepting automated track inspection (ATI) waivers. FRA's decisions on Class I carriers' ATI waiver petitions are the specific thing to watch — denial holds embodiment high, grant erodes it.
FAA rulemaking or 14 CFR Part 43/65 amendment explicitly requiring a certificated Airframe & Powerplant / Inspection Authorization holder to personally sign any airworthiness release where AI or automated NDT tooling generated the finding — i.e. barring machine-generated sign-off. Parallel: FMCSA codifying in 49 CFR 396 that CVSA out-of-service orders may only be issued by a certified CVSA-credentialed inspector, and that AI pre-screening cannot itself constitute an inspection. CVSA's North American Standard program already gates decal issuance to credentialed humans; formalizing the exclusion of automated issuance would harden it.
Post-incident precedent (NTSB findings after a rail derailment or hazmat release) assigning personal accountability to the signing inspector for accepting an automated defect classification without independent verification — as happened in scope-of-inspection findings following East Palestine. Named-individual liability in a consent order or a state criminal referral raises the cost of delegating the call and locks the ambiguity-owning function to a person.
The limit. trust_premium has no realistic route: the buyer of an inspection is a regulator or a carrier complying with one, not a consumer who can prefer a human, and no one pays extra for a human-signed brake check beyond what the rule requires. Also note the largest downside risk sits in embodiment, not capability — FRA granting broad automated track inspection waivers would cut both embodiment and headcount regardless of how the liability shield moves.
| New York-Newark-Jersey City, NY-NJ | 4,000 | $93,420 +1% |
| Dallas-Fort Worth-Arlington, TX | 890 | $126,010 +37% |
| Miami-Fort Lauderdale-West Palm Beach, FL | 820 | $100,100 +9% |
| Phoenix-Mesa-Chandler, AZ | 690 | $51,990 -44% |
| Chicago-Naperville-Elgin, IL-IN | 570 | $55,070 -40% |
| Los Angeles-Long Beach-Anaheim, CA | 560 | $101,040 +10% |
| Atlanta-Sandy Springs-Roswell, GA | 480 | $107,640 +17% |
| Houston-Pasadena-The Woodlands, TX | 480 | $92,570 +1% |
| Minneapolis-St. Paul-Bloomington, MN-WI | 180 | $140,880 +53% |
| Anchorage, AK | 110 | $131,060 +42% |
| Seattle-Tacoma-Bellevue, WA | 270 | $129,480 +41% |
We have no reported case of a named organisation automating this occupation. Not one deployment, not one announcement.
That is worth saying out loud next to a score of 67. The verdict above is about what the work exposes — what current AI could do to these tasks. It is not a claim that anyone has done it. For this occupation those two things have come apart completely: the capability argument is on this page, and the evidence column is empty.
Has AI actually changed your work? One tap, anonymous, and the running tally is public. Nothing else is asked of you.
Rather than check back: get the digest and we'll tell you what changed — or watch a single occupation from its own page.