How it is applied
You start from the observed failure and ask why it happened. The answer becomes the new starting point and you ask again. The process ends when the answer points to a condition that can be corrected in a lasting way, normally a process, a design or a maintenance criterion.
An example with a motor
- Why did the motor stop? Because the thermal protection tripped.
- Why did it trip? Because the winding overheated.
- Why did it overheat? Because the ventilation was blocked.
- Why was it blocked? Because dust built up on the fins.
- Why did it build up without anyone noticing? Because that motor is not on the cleaning program.
The corrective action lies in the last link: adding it to the program. Replacing the motor or resetting the protection would have left the cause intact.
Rules for making it work
- Each answer must be a verifiable fact, not an assumption. If it cannot be verified, it is recorded as a hypothesis and the evidence is sought.
- Ask about the process, not the person. “Because the operator was careless” is rarely a useful root cause.
- Follow each branch. If an answer has two possible causes, both are explored.
Limitations
The number five is a guideline: some chains end sooner and others need more steps. The technique is linear and can oversimplify failures with several concurrent causes; in those cases it is complemented by an Ishikawa diagram or a fault tree.
Its quality depends on who is asking: two people can reach different conclusions with the same questions. That is why it is best done as a team and backed up with data.