Escalating a failed plant inspection at an Australian workplace

A pre-use or scheduled inspection finds a problem with plant. The first decision is whether it can be used safely, not whether someone can close an inspection form. A good escalation process keeps the plant identifiable, prevents unintended use, assigns competent assessment and preserves the reason for return to service.

This guide is about workplace **plant**, such as machinery or equipment, not a road-vehicle defect system. Safe Work Australia's model Code of Practice for managing plant risks is the national model reference. Model codes do not have identical legal effect everywhere: check the applicable state or territory regulator and the plant's specific obligations before publishing local legal instructions.

Record the failed check and make the area safe

Identify the plant, location, inspector, date, observed condition and task for which it was about to be used. Describe the fault rather than only selecting “failed.” Include a photograph or test result where it helps a competent person assess the problem. If the condition may expose people to risk, prevent use through the workplace's isolation or lockout process and tell affected workers. The precise isolation method depends on the equipment and hazard.

Do not allow an inspector to choose a harmless-looking label merely because production is under pressure. A failed guard, unstable component or unexpected movement needs different immediate controls from an administrative document gap. If the defect could affect adjacent work, secure that area too.

Route the decision to the right person

Name the person who can assess the plant and decide what repair or further inspection is needed. That may be an internal competent maintenance worker, a specialist contractor, a manufacturer or an engineer, depending on the item and fault. A supervisor can coordinate work and restrict access without pretending to be the technical certifier.

The escalation record should distinguish observed facts from diagnosis. Write “hydraulic leak at hose connection” if that is what was seen; do not assert the cause until a competent assessment confirms it. Include the manufacturer's instructions and maintenance history where relevant. If an inspection certificate or registration requirement applies to a particular plant type, verify it with the jurisdictional regulator.

Track repair and controlled restart

Record the assessment, work performed, parts or tests, who completed them and any residual restriction. The person authorised under the workplace procedure should verify that required checks are complete before removing the hold. Brief operators on a changed condition. Safe Work Australia's model plant code discusses the need for communication and consultation with affected workers when restarting plant shut down for inspection, maintenance or cleaning.

Do not close the original failed item simply because a work order has been created. The inspection, repair, test and release should remain linked. If the plant is retired or moved to another site, record that disposition so an old “serviceable” status cannot reappear later.

Investigate repeated failures

Worked example: an intermittent guard fault

A pre-use check finds that a machine guard does not remain in the expected position. The operator records the observed behaviour and asset identifier, keeps the plant out of use under the local procedure and tells the supervisor. The supervisor arranges a competent assessment rather than resetting the inspection to “pass.” The assessor determines the cause, specifies repair and any testing needed, and records who may authorise restart. If the fault does not reproduce immediately, the original observation still matters; an intermittent issue is not resolved merely because a later check appears normal.

After repair, capture the test result and release decision in a record linked to the failed inspection. Brief the affected operators if the safe method or control has changed. If the same fault recurs, review the repair and operating conditions instead of creating another isolated closure. The precise technical checks depend on the plant and applicable local requirements; this example is a record and decision sequence, not engineering advice.

Questions before removing a hold

Has the person assessing the fault identified the specific plant and condition? Is there evidence of the repair or other control, and of a check that it worked? Are any use restrictions still in place? Do operators know the current status? Is the decision maker authorised under the workplace's actual procedure? If the answer to one of these is unknown, record the uncertainty and escalate it. A green software indicator alone is not an adequate release decision.

Repeated faults can point to unsuitable use, maintenance intervals, training, environment or design. Review trends and decide whether the risk controls need changing. A high count of completed inspections is not reassuring if the same serious defect recurs. Record what was changed and how the team will test that the change worked.

For software evaluation, see Complys AU WHS compliance software and ask for a demonstration of plant identity, failed check, hold and human release decision in the current product. This guide does not claim automatic severity classification, legal inspection certification or a plant lockout integration.

Where more than one contractor uses the plant, communicate the hold to each affected team. Confirm that the asset identifier on the repair record matches the failed item and that the authorised release is visible at the point of use. If the plant moves, transfer the restriction with it rather than leaving it in an old site file. Check that every team uses the same current asset status.