Onboarding a New Maintenance Tech: The First 90 Days

Why maintenance onboarding is different

A new operator learns one line. A new maintenance tech has to learn every machine in the building — plus where the shutoffs are, which breaker feeds what, how the plant wants work orders written, and forty unwritten rules like “the wrap machine says 60 PSI but runs best at 55.” In most small plants, none of that is written down. It lives in the heads of the senior techs, which is exactly why onboarding defaults to shadowing.

Shadowing works, but it's slow, it transfers habits along with knowledge, and it collapses when the senior tech is buried in breakdowns. A light structure — one page per month — makes expectations and progress easier to review. Here's the shape.

Days 1–10: Safety, systems, and the map

• Safety first, and specifically. Generic safety orientation isn't enough for maintenance. The tech needs your lockout/tagout procedures machine by machine, arc flash boundaries, confined spaces, and the location of every disconnect they'll ever throw. Don't sign this off until they've performed — not watched — a lockout on each major machine class.
• The utility map. Walk the plant tracing each system end to end: compressed air from compressor to drops, electrical from service entrance through panels, water, gas, dust collection. When something fails at 2 a.m., “which breaker?” can't wait for the senior tech's cell phone.
• The tools they'll log work in. Set them up in your maintenance system on day one and have them close their first work order — even a trivial one — in week one. Habits form immediately; if the first month is paper scraps, the system never sticks. (If your existing crew barely uses the system either, a new hire is a great excuse to fix that — see getting techs to actually use your maintenance software.)

Days 10–30: PM routes as a training course

Preventive maintenance is the best onboarding curriculum you own. PMs are scheduled (no 2 a.m. pressure), repetitive (the same machines come back around), and touch everything (a month of PM routes visits most of the plant). So put the new tech on PMs early and heavily.

The sequence that works: they shadow each PM once, they perform it observed once, then they own it. Good checklists are what make the third step safe — a PM written as “check conveyor” trains nobody, while one with specific steps and pass/fail readings teaches the machine while the work gets done. If your checklists can't train a new hire, they were never really protecting you from the old hires' turnover either; here's how to write ones that can.

By day 30, the tech should independently own the PM routes for your simpler assets, know the plant's layout cold, and be writing work order notes in your standard format.

Days 30–60: Corrective work, with a wingman

Month two adds repair work — the unpredictable half of the job. Structure it by escalation rather than shadowing: the new tech takes the work order first, does their own diagnosis, and calls for help when they hit their limit. They'll reach for help often at first; that's the point. Diagnosing for ten minutes and then getting unstuck teaches far more than watching someone else diagnose for ten minutes.

Two force multipliers for this phase:

• Work order history. Before touching a down machine, the tech reads its history. The machine may have failed the same way before, with the repair written down — if your team writes real completion notes. A new hire is the strongest argument for work orders that actually capture the fix, because they're the first person who benefits from every note the veterans left.
• Instant machine context. If your equipment carries QR code labels, the tech standing at an unfamiliar machine is one scan from its manuals, history, and open work — instead of one walk to the office and one interruption of a senior tech.

This is also the month to assign a small solo project — rebuild a spare pump, overhaul a non-critical conveyor section. A contained project builds confidence and shows you their diagnostic style while the stakes are low.

Days 60–90: Independence, and the payback

Month three, the training wheels come off: the tech takes normal work order assignments, joins the on-call rotation with backup, and owns their PM routes without check-ins. Your job shifts to calibration — reviewing their completed work orders weekly, watching for the two failure patterns of new techs (never asking for help, or never diagnosing before asking) and correcting early.

Then collect the payback. A new tech is the best documentation auditor you will ever have, because they can't fill gaps from memory. Every question they asked in 90 days is a gap: a PM step that assumed tribal knowledge, a machine quirk that lived in one person's head, a spare part nobody could find. Close out onboarding with a working session where their question list becomes checklist edits, asset notes, and labeled parts bins. Do that, and the next hire has clearer instructions and fewer undocumented steps — and your plant stops being one resignation away from losing how half its machines get fixed.

That's the real return on structured onboarding. The first 90 days aren't just about one tech getting up to speed; they're about converting your team's memory into a system anyone can be hired into.