How to write a work order that survives shift change

What a work order is for

A work order is the unit of owned maintenance work. It exists so a fault or planned task can move from “someone should look at this” to a completed record on an asset — with enough context that the next shift does not start from zero.

In a [CMMS](/guides/what-is-a-cmms), work orders are how plants:

• Capture reactive stops and planned PM as real jobs
• Assign ownership across technicians and contractors
• Preserve history for [downtime reduction](/guides/reduce-equipment-downtime) and reliability review
• Link parts and notes to the machine that needed them

If open work lives only in chat, you do not have a work-order process — you have a group thread.

Required fields that matter on the floor

Keep the form short. These fields earn their place:

• Asset — the specific machine or location (not “packaging area”)
• Problem / task statement — what is wrong or what planned work is due
• Impact / priority — why this matters to production or safety now
• Assignee / claim — who owns the next action
• Status and timestamps — reported, started, completed
• Parts / tools — known needs up front; actual usage on close
• Completion notes — what was found and what was done

Optional but high-value: photos, meter readings, related PM ID, safety permits. Do not add fields that only serve a monthly report if they slow the first report on the floor.

How much detail is enough

Enough detail means the next person can act without calling you.

Good enough usually includes:

• Symptom (noise, fault code, quality reject, no-start, leak)
• When it started / how often it recurs
• What was already tried
• Whether the line is down, restricted, or running with risk

Too little: “Line 2 broken.” Too much: a novel nobody will read under pressure.

Operators can open the work order with symptoms. Technicians refine diagnosis as they claim and complete. The handoff should add detail — not restart the story.

Priority from impact (not politics)

Priority should reflect production and safety impact, not who shouted last.

A simple plant-friendly scale:

1. Safety / environmental — stop and escalate
2. Line down / no production — immediate claim
3. Restricted rate / quality risk — same shift
4. Degraded but stable — plan into the next window
5. Improvement / non-urgent — backlog with a date

Write the impact in one line on the work order (“Sealer down — Line 2 stopped”) so priority is auditable later. Politics thrives when impact is invisible.

Parts, photos, and handoff notes

Evidence beats memory.

• Attach a photo of the fault, tag, or damaged part when safe
• List parts already known so someone can stage them
• On close, record parts actually used and the fix that worked
• Leave a handoff note if work paused mid-job (“waiting on belt; kit staged at cage B”)

This is what makes history useful the next time the same asset stops. Without it, every restore is a first restore.

Example work orders (good vs bad)

Bad — reactive

• Asset: (blank or “packaging”)
• Title: Machine broken
• Priority: High
• Notes: (empty)

Nobody can start intelligent work from that.

Good — reactive

• Asset: Packaging Line 2 — sealer
• Title: Sealer fault — intermittent jaw temperature drop; line stopped
• Impact: Line down; ~40 cpm lost since 14:10
• Detail: Fault appeared after changeover; temp swings 20°C; checked loose connection — still intermittent
• Priority: Line down
• Parts/tools: spare thermocouple on hand if needed
• Owner: claimed by Tech A

Bad — planned / PM

• Title: Do PM
• Checklist: 30 vague steps
• Window: “sometime this week”

Good — planned / PM

• Asset: Packaging Line 2 — sealer
• Title: Weekly sealer PM — jaw clean + thermocouple check
• Window: Friday changeover (25 min)
• Checklist: short, numbered, finishable in the window
• Parts: cleaning kit; spare thermocouple staged
• Owner: Packaging tech lead

How Corivo structures work orders

Corivo treats work orders as the floor system of record: faults and PM land on the asset, someone claims and starts, completion notes and parts stay with the history, and the next shift can continue without reconstructing a chat thread.

You do not need a perfect register to write the next good work order. Put the fault on the right machine, state the impact, assign an owner, and close with notes. That habit — repeated — is how plants turn maintenance into durable history instead of tribal memory.