■ HIGH RISK ■ Public Service & Government
Continuous sensor monitoring will hollow out the routine annual inspection, but code enforcement needs an accountable human signature — fewer inspectors, auditing data streams instead of riding car tops.
“IoT sensors monitor cables 24/7. Your annual visit was cute.”
Our AI replacement risk score — how we score jobs
An elevator inspector's rounds are a safety code made flesh: witnessing full-load safety tests, checking brake function and door protection, examining hoistway clearances and pit conditions, verifying firefighter-service operation, reviewing maintenance logs, and writing up violations that building owners must fix before the certificate gets issued. It's methodical checklist work built on deep code knowledge — ASME A17.1 is not a light read — and the authority to shut down a machine people ride every day.
The disruption is that the elevator now inspects itself, continuously. Major manufacturers ship IoT-connected units streaming door cycles, brake performance, ride quality, and fault codes to cloud platforms that predict failures before they happen — a data feed that makes an annual visual check look like diagnosing a patient from last year's photo. Remote monitoring platforms flag anomalies in real time; maintenance records are digital and auditable; and jurisdictions stretched thin on inspectors (most are) have every incentive to accept data-driven, risk-based inspection schedules that concentrate human visits where the data smells wrong.
But inspection is a legal act, not just a technical one. Certificates require an authorized inspector's judgment and signature; witnessing a safety test — watching the machine actually catch a falling car — is hard to replicate through telemetry that the maintenance company itself controls. There's an obvious conflict-of-interest problem in letting the manufacturer's sensors self-certify the manufacturer's maintenance, which regulators understand. The role compresses toward auditing: fewer routine visits, more data review, targeted physical inspections, and enforcement on the buildings whose numbers don't add up. A shrinking, more analytical corps — not an empty one.
Automatability: our editorial assessment of current and near-term AI capability
Pressure builds through 2030 as IoT-monitored elevators become the installed norm and jurisdictions adopt risk-based inspection schedules that trade routine annual visits for data review. Inspector shortages accelerate the shift — automation here fills vacancies as much as it cuts jobs. Witnessed safety tests, new-installation acceptance, and enforcement stay human beyond that horizon, because a certificate needs an accountable signature the sensor feed can't provide.
It's replacing the routine layer — continuous telemetry on brakes, doors, and ride quality beats an annual snapshot, and stretched jurisdictions are moving toward data-driven, risk-based schedules. But certificates require an independent human judgment, and letting the maintenance company's own sensors self-certify is a conflict of interest regulators won't fully accept. Inspections get fewer and more targeted, not extinct.
Surprisingly defensible in the near term — most jurisdictions face inspector shortages and an aging workforce, so demand for credentialed people is real today. Long-term, expect fewer positions doing more data auditing and fewer routine site visits. Enter with strong code knowledge and comfort with monitoring platforms, and you're positioned for the version that survives.
Plenty: hoistway housekeeping and pit conditions, unauthorized alterations, degraded components the sensor suite doesn't cover, and whether the maintenance log matches physical reality. Above all, sensors can't witness a safety test — verifying the machine actually arrests a loaded car — or exercise code judgment on unusual installations. That's the residual human core of the job.
Learn to read the telemetry platforms now, before your jurisdiction mandates it — auditing remote data will become a core competency. Keep your witnessed-testing and acceptance-inspection skills sharp, since those stay physical. And build expertise in the messy edge cases: altered, aging, and non-standard installations are where human judgment stays priced in.