Back to all posts
Plant Data7 min read·October 25, 2026

How Should You Structure Downtime Reason Codes?

TB
ThynkBlox Automation Team
Automation Engineering

To create downtime reason codes, start with a short hierarchy: a broad reason family followed by a specific reason that an operator can recognise. Define what each code includes, what it excludes and what to choose when the cause is uncertain. Keep planned work separate from the reason itself, and give one named owner responsibility for changes to the list.

What should a downtime reason code describe?

Use the code to answer one question: what prevented this equipment from producing during this interval?

Keep that separate from who should respond, whether the stop was planned and what a later investigation establishes as the root cause. Those are useful details, but they answer different questions.

For example, “waiting for material” describes the immediate production constraint. “Stores team” names a possible responder. “Poor replenishment planning” proposes an underlying cause. Do not make operators choose between these as though they were equivalent reasons.

Define the reporting boundary first. Is the record describing a machine, a production line or a whole process area? A machine waiting for its upstream neighbour needs a different description from the upstream machine that has actually failed.

For the broader reporting context, see what the OEE number hides. Agree what a stop represents before using its reason code to explain a performance figure.

How much hierarchy should you use?

Start with two levels: reason family → selectable reason. Add another level only where it supports a specific decision and operators can reliably distinguish the choices.

The following hierarchy is illustrative. Adapt the boundaries to your process rather than copying every entry.

Reason familySelectable reasonBoundary to define
EquipmentMechanical faultAn observed mechanical problem on the equipment being recorded
EquipmentElectrical or control faultAn observed electrical or control problem on that equipment
MaterialRequired material unavailableProduction is ready, but the required input is unavailable
FlowWaiting for upstream processThe recorded equipment cannot run because its upstream process is not supplying it
FlowDownstream unable to receiveThe recorded equipment cannot discharge into the next process
Production activityChangeoverWork required to switch between agreed production conditions
QualityAwaiting quality decisionProduction is held pending a defined quality decision
Maintenance activityPreventive maintenanceMaintenance work scheduled under the plant's agreed planning rule
UnconfirmedReason not yet establishedAvailable evidence does not support a more specific selection

Keep the selectable reasons meaningful. A code for every component is unnecessary unless that detail changes an action someone will take.

Use technical detail in the investigation record when the operator cannot establish it at the time of the stop.

What should each code definition contain?

Give each selectable reason a small definition card containing:

  • Label: the words the operator sees.
  • Use when: the observable condition that qualifies.
  • Do not use when: the closest competing condition.
  • Example: an illustrative situation.
  • Review owner: the role responsible for resolving disputed selections.

Consider “required material unavailable”. Its definition might say: use when the next required material is unavailable at the agreed supply point; do not use when material is present but cannot be processed because of an equipment fault.

Avoid labels such as “operational issue” or “production problem”. Rewrite them around an observable condition.

Validate the definitions with operators from different shifts. Give them the same sample stops and ask them to select codes independently. Where selections differ, examine the wording and boundaries before adding more choices.

How should operators classify an ambiguous stop?

Write a tie-breaking rule before the ambiguity occurs. A useful starting rule is to select the immediate constraint at the reporting boundary, using only what is known.

Illustrative example: a packing machine stops because no product arrives from upstream. The upstream mixer has a confirmed drive fault.

For the packing machine, select “waiting for upstream process”. For the mixer, select the relevant confirmed equipment reason. Do not describe the packing machine as having a drive fault.

If reporting later combines those records, define how related stops will be treated. Reason labels alone should not determine whether several equipment records represent one production loss.

When several constraints overlap, agree whether the record follows the first constraint or is split when the constraint changes. Document the choice and apply it consistently. Do not ask operators to invent a priority rule during each stop.

An alarm may provide evidence, but avoid treating every alarm as a reason code. Define which observed conditions justify a selection. The distinction is useful alongside our guide to alarm management.

How should planned work be classified?

Record the activity and its planning status as separate decisions.

“Changeover” describes an activity. “Planned” describes its relationship to an agreed schedule. The same distinction applies to cleaning, inspection and maintenance.

Define what qualifies as planned: which approved schedule counts, who approves it and when its status becomes fixed. Do not allow an unexpected stop to become planned simply because someone arranged work after production had stopped.

Also define the boundary between the activity and an interruption to it.

Illustrative example: a scheduled changeover includes tooling replacement and adjustment. A damaged clamp then prevents completion. Agree whether the interval changes to an equipment reason once the clamp becomes the constraint. Keep the original changeover context available.

Do not make exclusion from a performance calculation a property of the reason label. Document that reporting rule separately.

What should happen when the reason is unknown?

Provide an explicit “reason not yet established” choice. Do not require a confident diagnosis where none is available.

Separate uncertainty from missing review:

  • Reason not yet established: someone reviewed the event, but the evidence remains insufficient.
  • Awaiting review: the event has not yet been assessed.
  • Other defined reason: a known condition does not fit the current list and needs a description.

Give unresolved events an owner and an agreed review point. If evidence later supports a correction, retain the original selection and the reason for the change.

Do not force unknown events into a plausible equipment category to make a report look complete. Keep unresolved loss visible.

Review repeated unknown selections for a specific problem: unclear definitions, missing evidence or an unsuitable code list. “Other” should trigger that same review rather than become a permanent miscellaneous bucket.

Who should own changes to the code list?

Assign one accountable owner, with operations, maintenance and quality contributing where their boundaries are affected.

Require each proposed change to answer:

  • Which recurring situation cannot be classified clearly?
  • Why does an existing code not fit?
  • What decision will the new distinction support?
  • Can operators recognise it from available evidence?
  • How will historical records be compared after the change?

Prefer a clearer definition over a new code when that resolves the problem.

Keep code identifiers stable. If a meaning changes substantially, retire the old code and create a replacement. Record the mapping and the point at which the replacement applies; avoid silently rewriting historical meaning.

Include ownership, definitions and change rules in the agreed requirements. Our guide to what a functional design specification should contain explains where such decisions belong.

What common mistakes should you avoid?

Mixing causes, symptoms and departments. “Motor fault”, “machine stopped” and “maintenance” do not answer the same question. Choose a consistent level of description.

Making the list a parts catalogue. Add component detail only when it supports an agreed use and can be selected reliably.

Using reasons to assign blame. Describe the constraint first. Keep responsibility for corrective action separate.

Changing labels without reviewing their meaning. Check whether a rename preserves the definition or creates a different category.

Forcing one list onto unlike processes. Share broad families where useful, but allow process-specific reasons with clear mappings for comparison.

What questions come up before using the list?

How many downtime reason codes should we create?

Start with the smallest set that distinguishes the actions you need to take. Test it against representative stops. Add a code when a recurring, distinguishable condition supports a different decision.

Should operators select the root cause?

Ask operators to select the reason supported by current evidence. Record a confirmed root cause separately after investigation, and correct the initial reason only when the evidence warrants it.

Should every machine use identical codes?

Use common definitions for shared conditions. Add equipment-specific reasons where necessary, while preserving a clear mapping to the broader families.

When is a simple hierarchy insufficient?

Where one interval genuinely involves several constraints, keep a primary reason and related context rather than endlessly expanding the hierarchy. If safety-related decisions depend on the classification, develop those rules with a qualified assessor.

A useful final check is straightforward: can different operators classify the same example using the written definitions, and can each selection support a clear follow-up action?

Next step

Want to apply this to your plant?

Every project starts with a fixed-fee assessment, and you can stop after the report.