Five Whys is an investigation technique that chains questions about a problem until the team reaches causes and contributing factors that can be tested and controlled. The number five encourages depth without acting as a fixed target. An analysis may stop earlier, continue beyond the fifth question, or open more than one chain when the evidence points in different directions.
In industrial maintenance, the technique works best with a bounded problem and a causal sequence that can be examined. It helps organize hypotheses. Confirmation comes from records, inspection, measurement, and testing, because a convincing answer in a meeting can still be wrong.
What Five Whys delivers
The starting point is a precisely described event. Each answer to “why?” proposes a causal relationship that needs support. When the team finds more than one explanation compatible with the observed facts, the analysis opens branches instead of selecting the first account available.
The Five Whys guide from the Centers for Medicare & Medicaid Services recommends involving people who know the process, continuing beyond five questions when necessary, and validating a cause by asking whether the problem would probably recur after that cause was removed. The guide also warns that the technique may not identify every cause and recommends broadening the investigation with a cause-and-effect diagram when the problem requires a wider view.
The Institute for Healthcare Improvement reinforces another important limit: a problem may have multiple causes, and people who observe different parts of the system may build different answers. In industry, this is why operations, maintenance, engineering, and safety need to participate in relevant cases.
When to use the method and when to expand it
Five Whys is suitable when the event is well bounded, the team has access to the equipment and records, and the chain can be verified in the field. A protection trip, a component that failed earlier than expected, or an inspection that was not performed can be useful starting points.
Expand the investigation when there is:
- more than one failure mode capable of producing the same effect;
- interaction among equipment, process, environment, material, and human behavior;
- a safety, environmental, or high-consequence event;
- insufficient evidence to choose among competing hypotheses;
- recurrence after an action that should have removed the cause.
In these cases, a fishbone diagram helps open categories of factors, a fault tree makes logical combinations explicit, and barrier analysis checks why controls failed to prevent the event. The root cause analysis in maintenance guide organizes these tools within the same process.
How to apply Five Whys rigorously
- Describe the event. Record the equipment, affected function, date, shift, operating condition, protection that operated, and process effect.
- Preserve evidence. Separate alarms, trends, photographs, removed parts, work order history, the current maintenance plan, and accounts from people who observed the event.
- Form the first relationship. Answer why the effect occurred and state which evidence supports the answer.
- Test alternatives. Ask which other condition could produce the same effect and what would distinguish the hypotheses.
- Open branches when needed. A technical failure, a missing control, and an operating condition can coexist.
- Stop with a criterion. The chain is useful when the team finds a controllable cause, supported by evidence and connected to an action capable of reducing recurrence.
- Verify after the action. Define the indicator or inspection that will show whether the causal hypothesis was correct.
The cause may be physical, technical, human, or organizational. The method does not need to reach a management decision to be complete. It needs a testable explanation and a control proportional to the consequence.
Applied investigation: compressor trip on thermal protection
The case below is instructional. The facts, measurements, and causes were constructed to demonstrate the method and do not represent an actual customer event.
Event: air compressor CA-02 stopped at 2:37 p.m. while operating under load. The relay recorded a motor thermal protection trip. Air header pressure remained below the operating limit for 18 minutes.
Evidence preserved: relay log, oil and discharge temperature trends, a photograph of the cooler, temperature differential before and after the cooler, fan condition, the current maintenance plan, and the latest completed work orders.
Main chain: loss of heat transfer
- Why did the compressor stop? Thermal protection operated after temperature exceeded the configured limit. The relay log confirms the event.
- Why did temperature exceed the limit? The cooling circuit did not remove enough heat under that shift's load condition. The trend shows a continuous rise even with the fan running.
- Why was heat removal insufficient? Airflow through the cooler was restricted. Inspection found fins blocked by dust adhering to oily residue, and cleaning restored the expected differential during a controlled test.
- Why did the blockage reach that level? The current plan had no condition inspection or cleaning criterion for the cooler. Previous work orders covered oil and filters without recording fin condition.
- Why did the plan not control this condition? Plan review did not incorporate the particulate level and contamination observed in CA-02's actual environment. The baseline routine used during implementation remained unchanged.
This chain supports a causal hypothesis related to contamination control and maintenance plan review. It does not justify concluding that every compressor overheating event starts at the same point.
The branch that cannot be hidden
The oily residue on the fins opens another question: where did the oil come from? If a hose, seal, or separator is leaking, cleaning the cooler and increasing inspection frequency would address the consequence without removing the source that makes dust adhere. This branch requires a leak inspection and a comparison of oil consumption. If the evidence confirms a leak, it becomes an additional cause with its own action.
The fan must also be eliminated as a hypothesis. Current, speed, and rotation direction within specification help rule out a loss of active ventilation. Without these measurements, stating that blockage alone explains the trip would be premature.
From cause to verifiable action
The case calls for actions on different time horizons:
- containment: clean the cooler, verify the fan, inspect for leakage, and test the compressor under controlled conditions;
- correction: remove the oil source if confirmed and create an inspection with an objective blockage criterion;
- prevention: review the plan considering environment, load, and asset history, with an owner and interval approved by engineering;
- verification: monitor temperature, cooler differential, alarm recurrence, and fin condition during the next inspections.
The action enters the maintenance plan and becomes scheduled work. Verification needs to return to the asset history, because the absence of another trip does not by itself prove that the cause was removed. It may only mean that the operating condition has not recurred.
Errors that weaken the investigation
- Treating “human error” as a complete explanation. The action or omission needs to be described, and the context must be checked, including interface, procedure, training, workload, supervision, and existing barriers.
- Forcing a single chain. An effect may depend on several simultaneous conditions.
- Confusing correlation with cause. An item found after the event may be a consequence, a coincidence, or a contributing cause.
- Stopping at question five. The count does not replace evidence and control criteria.
- Closing without verification. An action without a follow-up measure leaves the causal hypothesis untested.
How Five Whys relates to FMEA
Five Whys starts from an observed event and deepens causal relationships. FMEA starts from functions and possible failure modes to anticipate risk and define actions before the event. The methods can share history and field knowledge, but they answer different questions.
After a confirmed analysis, the newly identified cause or failed control should feed the asset's FMEA guide in Portuguese. This keeps the learning from remaining isolated in the event report.
Frequently Asked Questions
Is it mandatory to ask exactly five questions?
No. Three questions may reach a verifiable cause in a simple case, while another case requires more than five or several chains. The criterion is the quality of the explanation and the evidence.
Does Five Whys always find the root cause?
No. The technique helps organize an investigation, but it can follow an incomplete chain, ignore parallel causes, or reflect the viewpoint of the people in the room. Complex cases require additional tools and field validation.
Must a root cause be organizational?
No. A cause may be physical, technical, human, or organizational. It must be supported by evidence, connected to the event, and addressable through an action capable of reducing recurrence or consequence.
How does the team know when to stop?
Stop when the team has a controllable and testable cause, knows which evidence supports it, defines an action proportional to risk, and establishes how the result will be verified. If another question only changes the level of description without changing a control or decision, the chain may already be sufficient.
When should a fishbone diagram be used?
Use it when the effect can arise from several categories of factors or when different participants produce incompatible chains. The fishbone diagram broadens the field of hypotheses, and Five Whys can deepen the branches that survive verification.
Bring the investigation into the maintenance workflow
An analysis only changes reliability when a confirmed cause becomes an action, a work order, a maintenance plan revision, and a follow-up check. The PM Run Advanced Study Group, GEA, holds invitation-only sessions focused on customers and partners. Customers and partners can request an invitation from their account executive. Companies that are not yet customers can contact the commercial team to learn about the initiative.
