What is root cause analysis in manufacturing?

Root cause analysis is the umbrella practice. It groups specific techniques such as 5-Why analysis, fishbone (Ishikawa) diagrams, fault tree analysis and failure mode and effects analysis (FMEA). Manufacturing quality and EHS teams use RCA at the analysis step of any structured problem-solving process, from a CAPA cycle to an 8D investigation.

1How does root cause analysis work?

A team begins with a clearly defined problem statement supported by data: what the deviation is, when and where it occurred, how often, and what the measurable effect is. Investigators then choose a technique that fits the problem class. The chosen technique produces a documented causal chain or diagram that the team can challenge with evidence.

The output of RCA is not just a ’cause’ but a verifiable proposition: if this cause is removed, the symptom should not recur. The corrective and preventive actions that follow target the validated cause, and recurrence detection over the following weeks confirms whether the analysis was correct.

2Why does root cause analysis matter for manufacturing?

Without RCA, teams react to symptoms and the same non-conformities come back. RCA is the analysis step that separates a closed-loop improvement system from a firefighting one.

ISO 9001:2015 clause 10.2 requires organisations to determine the causes of a non-conformity before taking corrective action, which makes RCA a regulatory expectation in any certified quality management system. The American Society for Quality identifies root cause analysis as one of the foundations of quality engineering and lists it alongside statistical process control and design of experiments as a core discipline.

3Techniques used in root cause analysis

Each technique fits a specific class of problem. Manufacturing teams typically learn three or four and apply the one that best matches the situation:

1. 5-Why analysis. Iterative questioning to drill from symptom to underlying cause. Best for single-thread causal chains with clear cause-and-effect linkage.
2. Fishbone (Ishikawa) diagram. Categorises possible causes across people, methods, machines, materials, measurement and environment. Best when many candidate causes exist and need to be sorted before testing.
3. Fault tree analysis. Top-down logical tree linking the failure to the combinations of events that could produce it. Best for safety-critical or high-consequence failures.
4. Failure mode and effects analysis (FMEA). Proactive technique that ranks potential failure modes by severity, occurrence and detection. Best for preventive analysis on a new product or process.
5. Pareto analysis. Ranks the contributing causes by frequency or cost to focus action on the most consequential ones.