Why Fixing It Twice Means You Didn’t Fix It the First Time

Every repeat failure has two price tags. The first is the repair cost and downtime that show up on the maintenance budget. The second is the implicit confirmation that the previous repair didn’t solve anything. It just reset the clock. Repeat failures aren’t a parts problem or a maintenance team problem. They’re a diagnosis problem. The fix is asking a different question before the machine goes back online.

Why the root cause doesn’t get found

Production pressure to restore operation quickly leaves no time for investigation. Symptom-level repairs feel complete. The bearing was replaced, the machine is running, and the work order is closed. But the failure investigation requires evidence that’s usually gone before anyone thinks to collect it: the failed component discarded, the fault log cleared, and operating conditions at the time of failure unrecorded.

Put simply, most CMMS work orders capture what was done, not why the failure occurred.

When a formal investigation is warranted

Not every failure justifies a structured root cause investigation. The analytical investment should match the consequence and recurrence risk. RCFA is warranted when:

• A critical asset failure caused significant unplanned downtime
• The same failure recurs on the same asset within a defined window
• The repair cost is unusually high relative to the asset’s normal maintenance profile
• The failure occurred despite a PM task specifically designed to prevent it

For non-critical assets with low failure consequences and no recurrence pattern, a reactive repair can be a rational choice. The mistake is applying that same logic to assets where the cost of repeat failure is real.

What a practical investigation looks like

The two most broadly applicable methods don’t require specialized software or a reliability engineering background:

• 5 Why analysis traces a single causal chain by asking “why” repeatedly until the underlying condition is identified. For example, a pump overheats, the cooling fan is clogged, the air filter was never replaced, and it wasn’t on the PM schedule. The root cause is a gap in the maintenance program, not the fan. This method works well for failures with a clear sequence and a single causal path.
• Fishbone analysis organizes potential causes across six categories — machine, method, material, manpower, measurement, and environment — and prevents teams from fixating on the most obvious cause while systemic contributors go unexamined. It’s better suited to failures where multiple contributing factors are suspected.

Both methods require gathering evidence before the machine restarts. The failed component, the fault log, and the operating conditions at time of failure are the inputs. Without them, the investigation is speculation.

What changes when investigation becomes a habit

Plants that apply structured failure investigation consistently show 30-50% fewer repeat failures than those that don’t, per maintenance benchmarking research. The CMMS failure history becomes a genuine diagnostic resource rather than a completed work order log. Patterns emerge across assets. The same failure mode on multiple similar machines points to a systemic cause that individual component repairs will never address. A corrective action without an assigned owner and a verification step isn’t a corrective action but a note.

The repeat failure cycle is expensive

There’s the repair cost, the downtime, and the carrying cost of a maintenance program that keeps solving the same problem. Root cause investigation doesn’t require a specialist. It requires asking why before the failed component goes in the bin. The answer is usually simpler than expected. The habit of asking is what’s hard to build.