5 Whys Root Cause Analysis for Maintenance
A step-by-step method, fictional pump-failure example, causal evidence, common traps, and criteria for moving to full RCA.
Read 5 Whys guideA structured technical path for maintenance teams, from 5 Whys and fishbone investigations through FMEA, P-F curves, RCM task selection, fault-tree logic, and Weibull analysis. Every method connects evidence to the decision it can legitimately support.
5 Whys, fishbone analysis and RCA investigate actual failures. Fault trees model combinations that can defeat a system function. FMEA identifies potential failure modes; FMECA adds criticality assessment. P-F analysis examines detection and response, Weibull analysis models failure-age behavior, and RCM converts validated functional and consequence evidence into a technically applicable maintenance policy. These are connected methods, not interchangeable names for the same exercise.
RCM is a decision process, not a target PM percentage. A useful analysis can result in condition monitoring, age-based work, authorized failure-finding, redesign, or deliberate corrective maintenance. Safety, engineering, OEM and regulatory controls still govern what is allowed.
Use observed failure evidence to reconstruct why the event occurred. Do not assume a neat story is a proven causal chain.
A step-by-step method, fictional pump-failure example, causal evidence, common traps, and criteria for moving to full RCA.
Read 5 Whys guideBroaden the analysis for consequential or complex events: multiple causes, causal testing, action ownership and effectiveness verification.
Read RCA guideExplore interacting equipment, process, measurement and work-system hypotheses; verify each branch before selecting corrective action.
Read guideMake the failure register useful enough to guide controls, rather than simply accumulating scores.
Structure function, functional failure, failure mode, effects, consequences, existing controls and decision ownership.
Read FMEA guideCompare FMEA and FMECA, learn consequence-first ranking, and avoid mistaking an RPN for actual probability.
Read FMECA guideFollow a fictional conveyor analysis from required function to failure-mode register, credible controls and CMMS implementation.
Read guideUse consequence and technical applicability to determine which policy earns its place in the maintenance program.
Understand functional context, failure effects, consequences, and the range of legitimate failure-management policies.
Read RCM overviewMove from failure analysis through technical task selection, failure-finding, redesign, and a controlled review cycle.
Read decision-logic guideAssess warning-time uncertainty, actual monitoring coverage, corrective lead times and whether intervention can happen before F.
Read guideUse credible failure behavior and consequence to choose effective policies and executable PM or engineering actions.
Read guideBuild a feedback loop through controlled PM tasks, corrective work, failure reporting, and effectiveness review.
Translate the selected failure policy into tasks, triggers, and work that can actually be executed.
Read strategy guideTrack failures, actions, verification and updates back into the maintenance program.
Read FRACAS guideUse fault trees to find combinations that can defeat system functions, and Weibull modeling to test life-distribution hypotheses. These approaches require explicit boundaries, credible data, and uncertainty.
Model redundant cooling systems, common-cause risk, cut sets and the assumptions required for defensible quantitative results.
Read guideInterpret beta and eta, include censored survivors, and challenge whether a time-based replacement can actually help.
Read guideThese educational articles reference SAE JA1011 (2024) for RCM process evaluation, IEC 60812:2018 for FMEA and FMECA, and ASQ for Five Whys, ASQ Fishbone for cause-and-effect investigations, IEC 61025:2006 for fault-tree analysis, and the NIST Engineering Statistics Handbook for Weibull model definitions. They are practical educational guidance, not a replacement for purchased standards, engineering review, safety rules, or controlled plant procedures.