Downtime numbers are useful only when everyone measures them the same way.
The practical test is simple: could a person who was not there understand what happened and make a better decision because the record exists? Does downtime start when the machine stops, when maintenance is called, or when production declares the loss?
Chapter focus: Turn this topic into a repeatable logging habit that survives shift pressure and creates usable history.
Define start and end
Does downtime start when the machine stops, when maintenance is called, or when production declares the loss? When this is not documented well, the maintenance team often pays twice: once during the original breakdown and again when the same failure returns.
Imagine that Different definitions produce arguments that no report can solve. The response is to Write the definition and use it consistently. The objective is not a diary of every movement; it is enough evidence to understand the state, the reasoning, and the decision that followed.
Separate response, diagnosis, repair, and waiting
A twelve-hour event may include ten hours waiting for a part and one hour of actual repair. Good maintenance records are designed around reproducibility. Another competent person should be able to understand what was observed, what was tested, and why the final action made sense.
A useful example is this: This distinction changes the improvement action. Make the rule explicit: Use simple timestamps or duration categories. Once the team applies the same rule consistently, search, handover, and reliability analysis become much easier.
Practical check: Look at three recent records related to separate response, diagnosis, repair, and waiting. Can a different technician understand the facts without asking the original author? If not, identify the one missing field or wording rule that would fix the problem.
Distinguish maintenance loss from production loss
Not every minute that a machine is stopped is caused by maintenance. This matters because maintenance work is full of interruptions, shift changes, and incomplete information. A useful record should reduce uncertainty for the next person, not merely prove that somebody typed something into a box.
For example, Material shortage, quality hold, cleaning, changeover, and operator availability may overlap. The practical standard is: Agree on how mixed events are classified. Preserve the evidence that changed the diagnosis and keep the wording specific enough to search later. Precision is more valuable than length.
Field Example
Suppose a air compressor stops with a low air pressure alarm. The weak record says only that the machine stopped and was reset. A useful record identifies the asset, captures the first alarm or physical symptom, states the decisive checks, records the confirmed or suspected cause, describes the action, and explains how normal operation was verified. If the cause is not known, the record lists what was ruled out and leaves a visible follow-up rather than inventing certainty.
This style of entry pays for itself when the event returns. The next technician can compare the new symptom with the previous evidence, confirm whether conditions match, and skip checks that were already shown to be irrelevant. For supervisors and engineers, the same structured fields make the event countable and comparable. One short, disciplined record therefore supports troubleshooting, handover, analysis, and improvement at the same time.
Chapter Action Checklist
• Review the current practice related to define start and end.
• Choose one wording or field standard from this chapter and pilot it this week.
• Find one recent record that would have been easier to troubleshoot with better evidence.
• Agree how this information will be reviewed or analyzed, not merely stored.
Figure. Timestamps let the team distinguish response, diagnosis, repair, and restoration.