How is a fishbone diagram used in maintenance root cause analysis?
A fishbone diagram is a cause-and-effect tool that groups plausible contributors to a clearly defined failure event. Maintenance teams use it to explore interacting equipment, operating, materials, methods, measurement and environmental factors, then test candidate causes against physical and recorded evidence. The diagram generates hypotheses; it does not independently establish root cause or prove which countermeasure will work.
Use fishbone analysis when the failure could have more than one contributor.
A single narrow causal sequence may be suitable for 5 Whys. Fishbone is useful when the evidence points toward interacting technical and organizational factors, or the team is starting to anchor on the first visible component failure. It gives less vocal participants a structured way to suggest alternate explanations.
Write an effect statement with an asset tag, exact failed function, event boundary and operating condition. 'Conveyor failed' is weak. 'Transfer conveyor CV-21 stopped moving under normal commanded operation at 14:06 due to a motor overcurrent trip' is specific enough to investigate. Preserve competing facts such as a recent gearbox change, product load, temperature and alarm sequence.
Make cause categories prompts for investigation, not predetermined blame.
Common 6M categories are Machine, Method, Material, Measurement, Manpower or People, and Mother Nature or Environment. Industrial teams may substitute Design, Controls, Maintenance Execution, Supply Chain and Operations if those headings produce stronger examination.
Do not count proposed causes in each category as a measure of risk. A long branch of guesses with no evidence is less meaningful than a single confirmed design weakness. Allow a contributor to appear in several categories where its interaction matters.
| Branch | Plausible candidate | Evidence that could support or refute it |
|---|---|---|
| Machine / design | Motor bearing drag or undersized drive | As-found bearing condition, nameplate, design load and trip settings |
| Method / execution | Incorrect coupling alignment after repair | As-left measurement, acceptance criteria and repair records |
| Material | Product accumulation or change in loading | Lot data, throughput rates and observed load profile |
| Measurement / controls | Current sensor error or nuisance trip | Independent measurement, calibrated instrument and event logs |
| People / work system | Unclear fault-response procedure | Current instruction, training and actual shift practices |
| Environment | Temperature or moisture intrusion | Ambient records, enclosure condition and ingress evidence |
Run the investigation as a transparent hypothesis review.
Invite people closest to the event, including operators, the technician who saw the as-found state, a controls specialist and a planner who can retrieve job history. Start with silent cause proposals before group discussion to reduce authority bias. Draw first-level causes, add subcauses and mark each branch as observed, plausible, contradicted, or untested.
For each high-value candidate, identify a test, the owner, and any preconditions for safe inspection. Freeze and preserve relevant alarms, photos and failed parts before corrective work destroys evidence. Be explicit when production has already restarted and crucial evidence has been lost.
- Confirm one event statement and operating boundary.
- Generate multiple technically plausible explanations.
- Link claims to CMMS history, measurements, physical findings, operating context and engineering records.
- Prioritize discriminating tests that could actually rule causes in or out.
- Escalate unresolved high-consequence branches to broader structured RCA.
- Turn verified contributors into controlled actions and a measured review.
Example: a packaging conveyor trips after a gearbox replacement.
Fictional event: CV-21 trips three times on motor overcurrent following a gearbox change. The work order shows a completed replacement, but contains no as-left alignment measurement. Operators report heavier product accumulation during the affected shift. Those observations create two candidate branches: mechanical resistance after the change and load accumulation.
A properly supported conclusion requires checking actual current trend, coupling alignment, gearbox rotation resistance, product flow and overload settings. Discovering an absent alignment measurement demonstrates a documentation or verification gap; it does not by itself prove the alignment caused the motor trip.
If measurements confirm both elevated friction and excess accumulation, preserve a two-branch causal explanation. Separate immediate containment such as clearing accumulation from permanent controls such as engineering-accepted alignment checks or process-flow changes.
Do not repeatedly restart tripped or hot equipment to gather data. Use site-approved troubleshooting, isolation, electrical safety and engineering authority. A diagram is never authorization for unsafe testing.
Close the cause loop, not the meeting.
An actionable output records the suspected cause, actual test result, confidence or uncertainty, selected countermeasure, change authorization, action owner, CMMS reference and exposure-based effectiveness criterion. Effectiveness might be normalized over conveyor operating hours and similar product load, rather than one uneventful shift.
If the pattern returns, reopen branches before adding another PM task. Feed confirmed mechanisms into FMEA or the RCM policy record to prevent the same blind spot from recurring elsewhere.
What should the final fishbone deliver?
A ranked and evidence-labeled set of competing mechanisms, not an attractive one-page claim of root cause. Convert verified contributors into engineered or work-system controls and review recurrence against comparable operating exposure.
- Preserve uncertainty and multiple causal branches.
- Differentiate observations from hypotheses.
- Ensure physical tests are authorized and safe.
- Verify corrective effectiveness, not merely action closure.
Technical references
- ASQ: Fishbone (Ishikawa) Diagram — Explains cause-and-effect brainstorming, standard cause categories and the need for follow-up analysis.
Move from the whiteboard to evidence.
Use the verified cause record to improve equipment, procedures and maintenance policy.
RCM learning path See facility-specific XRVE services