CRITICAL RISK ■ Public Service & Government

Will AI Replace Meter Reader?

Yes — this one barely needs AI. Smart meters report themselves, utilities have spent two decades installing them by the tens of millions, and each installation permanently deletes a walking route. The job survives only where the rollout hasn't finished.

94%

Smart meters don't need anyone to squint at tiny numbers.

Our AI replacement risk score — how we score jobs

Why Meter Reader scores 94%

Meter reading is honest, physical, repetitive work: walk or drive a route of hundreds of stops, find the meter (behind the hedge, under the porch, guarded by a dog with opinions), record the reading accurately, note damage or signs of tampering, and feed it all back for billing. Rain, heat, and the occasional hostile homeowner included. Accuracy matters because every misread becomes a billing dispute; access matters because a locked gate means an estimated bill and an annoyed customer.

The technology that kills the route isn't fancy. Advanced metering infrastructure — smart meters — transmits usage automatically, continuously, with no human within miles. Utilities across the US, UK, and EU have been deploying them at massive scale since the 2000s, driven less by labor savings than by grid management: time-of-use pricing, outage detection, and demand forecasting all require the data streams smart meters provide. Even the halfway technologies — drive-by radio reading, where one van harvests a day's walking routes in an hour — decimated headcount before full smart deployment. Every meter swapped is a stop that never needs visiting again, and the swap only happens once, in one direction.

The residual work: territories where rollout lags (rural areas, older gas and water systems, jurisdictions where customers can refuse smart meters), periodic physical inspections for safety and tampering, and final reads on move-outs where the old hardware persists. Water utilities trail electric, so some routes will outlast others. But this is a rearguard, and utilities typically manage the decline humanely through attrition — retiring readers aren't replaced; remaining ones become field-service techs. Our 94 isn't a prediction so much as a progress bar.

Which Meter Reader tasks can AI automate?

Walking or driving routes to record meter readingsHIGH
Entering readings into handheld devices for billingHIGH
Inspecting meters for damage, leaks, or tamperingMEDIUM
Gaining access to meters on private property and managing dogs and locked gatesMEDIUM
Performing final reads and service verifications at move-outsHIGH
Reporting hazards and service issues found in the fieldLOW

Automatability: our editorial assessment of current and near-term AI capability

When will it happen?

This is late-stage automation: smart meter penetration is already high across electric utilities in most developed markets, and the remaining routes — rural areas, water systems, opt-out customers — are a shrinking tail being cleared through the late 2020s and into the 2030s. Utilities manage the wind-down by attrition, so mass layoffs are rare; the job simply stops being hired for. If you're reading meters today, you're working the last chapters.

How to stay ahead

  • 01Transition to field service technician — utilities retrain readers to install and maintain the smart meters replacing them, often with a pay bump.
  • 02Pick up utility-relevant certifications: electrical safety, gas fitting, water distribution — the trades side of utilities is short-staffed.
  • 03Aim at line locating, leak detection, or damage prevention roles, which keep you outdoors and in demand.
  • 04Use utility tuition benefits now; most companies signal reader phase-outs years ahead, and the retraining window is open.

Meter Reader & AI: common questions

Do meter readers still exist, and for how long?

They exist wherever smart meter rollout is incomplete — parts of rural America, many water utilities, regions with customer opt-outs, and countries earlier in deployment. But every route shrinks as installations proceed, and utilities almost never hire new readers; they let the role wind down through retirement. Realistically the occupation persists in pockets into the 2030s while ceasing to be something anyone enters.

Was it AI that killed meter reading?

No — radio and cellular transmitters did, and they're older than the AI boom. Smart meters report usage automatically; drive-by radio systems before them let one van replace a dozen walkers. AI enters only at the margins, like image recognition for photo-verified reads and analytics that flag tampering. Meter reading is a reminder that most job automation is plumbing, not intelligence.

What do utilities do with meter readers when routes disappear?

The better ones retrain them. Reading routes wound down slowly and predictably, so utilities commonly move readers into meter installation, field service, leak detection, or line locating — jobs that pay more and are chronically short of people. If your employer offers apprenticeships or tuition support, that's the intended exit ramp. The workers stranded are usually contractors, whose agencies simply lose the contract.

Is any part of the job automation-proof?

The eyes-on-the-ground part, somewhat. Smart meters report their own data but not the corroded pipe beside them, the meter box someone's bypassed, or the gas smell in the basement stairwell. Utilities still send humans for safety inspections, tamper investigations, and service work — but those are technician jobs with different training, not reading routes. The observation skill transfers; the occupation doesn't.

Related jobs