Failure analysis: from symptom to root cause (RCA)

Failure analysis, or root cause analysis (RCA), is a structured method for finding the root cause of a failure instead of treating the symptom. You don't just repair what broke; you work out why it broke, so the same failure doesn't keep coming back. Root cause analysis moves your maintenance from reacting to preventing.

Symptom versus cause

The difference between symptom and cause determines whether you truly solve a failure or merely postpone it. The symptom is what you observe: the machine stops, a pump leaks, a bearing runs hot. The root cause is the underlying condition that produces that symptom.

Take a bearing that fails after three months. It's tempting to conclude it was "just a broken bearing", replace it and move on. But when you analyse the failure, it turns out the bearing fails because of a misalignment between the motor and the pump. The resulting vibration loads the bearing far beyond its design. Replace only the bearing and the alignment is still wrong, so the new bearing fails within a few months too. Correct the alignment and the failure is gone for good.

So the bearing was the symptom; the misalignment was the root cause. RCA forces you to make that distinction before you choose a fix.

Method 1: the 5 Whys

The 5 Whys (why-why analysis) is the simplest RCA method. You start at the symptom and keep asking "why?" of the previous answer until you reach a cause you can actually eliminate. Five is a guideline, not a rule: sometimes you're there after three steps, sometimes after seven.

A worked example on the bearing from the previous section:

1. 1Why did the pump stop? The bearing failed.
2. 2Why did the bearing fail? It was running with excessive vibration.
3. 3Why was there excessive vibration? The motor and pump were not aligned.
4. 4Why was the alignment wrong? The last assembly wasn't aligned to procedure.
5. 5Why didn't that happen? The assembly procedure required no alignment check and the tooling was missing.

The root cause here isn't the bearing, but a missing step in the work instruction and the absence of alignment tooling. That is the level at which you take action: complete the procedure, make the tooling available, instruct technicians. Stop too early, for example at "excessive vibration", and you're treating a symptom again.

The strength of the 5 Whys is speed: you need no software or forms, just the right people around the table. Its limitation is that it follows a single line of causation. For failures with several concurrent causes, that single line falls short. Then a broader tool is called for.

Method 2: the Ishikawa diagram

Where the 5 Whys follows one line, the Ishikawa diagram (also called the fishbone or cause-and-effect diagram) shows several possible causes at once. You place the problem as the "head" of the fish on the right, draw a horizontal spine, and let diagonal "bones" branch off it, each representing a cause category. Along each bone, you gather every conceivable cause with the team.

In maintenance and production you usually use the six Ms as categories:

• Man — knowledge, experience, instruction, human error. For example: technician not trained on alignment.
• Machine — the installation itself: wear, design, age, wrong settings.
• Method — working practices and procedures: missing or unclear work instruction, wrong assembly sequence.
• Material — parts and raw materials: wrong or deviating component, poor-quality lubricant.
• Measurement — measuring instruments and inspection: uncalibrated instrument, no vibration measurement, missing check.
• Milieu (environment) — the surroundings: temperature, humidity, dust, vibration from neighbouring machines.

By working through each category systematically, you force yourself to look wider than the first, most obvious cause. The fishbone is a brainstorming and structuring tool: it shows you where the candidate causes sit. Which cause was actually decisive, you then confirm with data from your failure history, measurements, or a targeted 5 Whys per suspect branch. The two methods reinforce each other: the Ishikawa diagram broadens, the 5 Whys deepens.

For a clear background on the layout and categories of the fishbone diagram, see the ASQ explanation of the fishbone diagram.

When to use RCA

A full failure analysis takes time. You therefore don't run it on every minor failure, but focus the effort on the cases where it pays off:

• Recurring failures — the same failure keeps returning, a sign that you've been treating symptoms so far. This is exactly what RCA prevents.
• Critical failures — a failure with major consequences for safety, production or cost, even if it occurs only once.
• Failures of unknown cause — when it isn't clear what happened and recurrence is realistic.

For one-off, insignificant failures, repairing and recording is enough. The recording itself matters, though: without reliable failure history you won't spot recurring patterns and you'll miss precisely the failures that deserve an analysis. Metrics such as MTBF and MTTR help make those patterns visible.

From root cause to prevention

An RCA is only finished when the root cause leads to a concrete measure that you anchor in your maintenance plan. The outcome feeds your prevention in several ways:

• Adjusting the maintenance plan — a new or revised task, for example a periodic alignment check or vibration measurement.
• Adjusting procedures — adding a missing step to the work instruction or assembly procedure.
• Improving the design or component — a different bearing type, better seal or more robust setup.
• Feeding your risk analysis — repeated root causes are valuable input for a structured reliability and FMECA approach, in which you systematically assess failure modes and their effects.

That closes the loop: a failure leads to an analysis, the analysis to a measure, and the measure prevents recurrence. If you want to take the step towards systematic reliability analysis, read on about what FMECA is.