Root cause analysis (RCA) is a structured method for finding the underlying cause of a failure or defect, so the corrective action prevents recurrence instead of only treating the symptom.
RCA usually follows the same steps. Define the problem precisely: what, where, when, how much. Build a timeline from normal operation to the event. Separate root causes from contributing factors and symptoms. Choose corrective actions, then verify later that they worked.
Common tools include the 5 Whys, which asks why repeatedly until reaching a controllable cause; the Ishikawa or fishbone diagram, which sorts candidate causes into categories such as man, machine, method, material, measurement and environment; fault tree analysis, which maps logical combinations of failures; and Pareto analysis, which ranks causes by frequency or impact. Formal methods such as 8D or A3 wrap these tools in a team process.
Limits are well known. The 5 Whys and fishbone are easy to run but not rigorous for complex events, and they tend to stop at the first plausible answer or at human error. Many failures have several interacting causes. RCA also depends on data: if the historian did not record the relevant signal, the analysis becomes guesswork. AI can help by searching large datasets and past incident reports for correlated events, but correlation is not cause, and the team still has to confirm it.
Common root cause analysis tools
| Tool | What it does |
|---|---|
| 5 Whys | Asks why repeatedly until reaching a controllable cause |
| Ishikawa (fishbone) diagram | Sorts candidate causes into categories such as machine, method, material |
| Fault tree analysis | Maps logical combinations of failures |
| Pareto analysis | Ranks causes by frequency or impact |
Questions about Root cause analysis
What are the steps of a root cause analysis?
RCA usually follows the same steps. Define the problem precisely: what, where, when, how much. Build a timeline from normal operation to the event. Separate root causes from contributing factors and symptoms. Choose corrective actions, then verify later that they worked.
When is the 5 Whys method not enough?
For complex events. The 5 Whys and the fishbone diagram are easy to run but not rigorous, and they tend to stop at the first plausible answer or at human error. Many failures have several interacting causes.
Can AI do root cause analysis?
Not on its own. AI can search large datasets and past incident reports for correlated events, but correlation is not cause, and the team still has to confirm the finding. If the historian did not record the relevant signal, the analysis becomes guesswork.
Root cause analysis in real plants
- Toyota Industries and Sight Machine: Toyota Industries uses Sight Machine on Microsoft Azure at its Nagakusa Plant to unify paint shop data and find the drivers of paint defects in bumper production.
- ANDRITZ and IndustrialMind.ai: ANDRITZ selected IndustrialMind.ai to deploy AI agents for drawing review, BOM generation and root cause analysis in hydraulic equipment parts manufacturing; a pilot phase is moving to a broader rollout.
- Daikin and Hitachi: Daikin and Hitachi began a trial of a generative AI agent for equipment fault diagnosis at Daikin's Rinkai commercial air conditioner factory in Sakai, Osaka.
- Parkland and Seeq: Parkland's Burnaby refinery uses Seeq for daily monitoring and incident investigation of its main air blower, sitewide inferential modeling, boilers, wet gas compressor and flare stack.
- Saint-Gobain Weber and Braincube: Saint-Gobain Weber uses Braincube Real-Time Process Optimization across 9 production sites, including on its bagging process to limit overfilling.
Vendors linked to Root cause analysis
Augury (2) · Braincube (1) · Hitachi (1) · Seeq (1) · Sight Machine (1) · Instrumental
Related terms
Statistical process control (SPC)
Statistical process control (SPC) is the use of statistical methods, mainly control charts, to monitor a production process and detect when its variation goes beyond what its normal, stable behavior would produce.
Anomaly detection
Anomaly detection is the identification of observations or patterns that deviate significantly from normal behavior in data.
Condition monitoring
Condition monitoring is the ongoing measurement of machine parameters such as vibration, temperature, or oil quality to detect significant changes that indicate a developing fault.
Prescriptive maintenance
Prescriptive maintenance is a maintenance approach that goes beyond predicting when equipment will fail and recommends what action to take, when, and with which resources to prevent or limit the failure.
CMMS (computerized maintenance management system)
A CMMS (computerized maintenance management system) is software that keeps a database of a site's assets and maintenance work.