How many hours did the equipment actually operate before failing?

Define what counts as operation and failure

An injection moulding machine may remain energised while waiting for material; a compressor may run without delivering the required service. Reliability analysis requires a definition of effective operation and of functional failure.

Depending on the asset, exposure can be expressed as actual production hours, hours under load, cycles, starts or time in a defined condition. Comparisons between machines require consistent units and criteria.

• Asset identification and its required function.
• States that count as operation and conditions to be excluded.
• Functional failure criteria agreed with maintenance.
• Start and end of downtime, return to service and subsequent exposure.
• Signals or counters actually included in the instrumentation.

The CMMS and operational history capture complementary information

A CMMS manages work orders, technicians, procedures, spare parts, costs and intervention closure. KERN IoP can provide actual operating hours, cycles, states, alarms and physical variables before or after the event, depending on the available signals.

Relating the two histories requires agreed identifiers and time references. Automatic integration requires verification of each system’s connectivity and scope; monitoring alone does not establish that integration.

• Before: accumulated exposure, conditions, alarms and trends.
• During: event, diagnosis, intervention and waiting periods.
• After: return to production, behaviour under defined exposure and recurrence.

MTBF and MTTR require consistent definitions

For a population of repairable assets and comparable functional failures, a descriptive calculation is MTBF = actual operating hours ÷ number of observed functional failures. If 840 operating hours and four qualifying failures were recorded over a period, the observed MTBF would be 210 hours. This describes the observed average; it does not predict the timing of the next failure.

When calculating active repair time, MTTR = sum of active repair times ÷ number of repairs. State how response time, diagnosis, waiting for spare parts, testing and release are treated. Total downtime may differ from active repair time.

When comparing equipment or periods, document the population, failure criteria and actual accumulated exposure. The example values are illustrative.

How to verify the result of an intervention

Closing a work order confirms that an activity was carried out. Assessing its effect requires subsequent observation of the equipment under known conditions. Historical data allows recurrence, event duration and signal trends to be reviewed where those data are available.

A brief period without another failure is insufficient to establish that the cause has been eliminated. Record the operating hours or cycles since the intervention and the observed operating conditions.

How to reconstruct downtime using machine and maintenance data

Continue exploring

• The maintenance management page
• How to incorporate signals and operating records
• How to access machine histories

Editorial basis: KERN IoP’s 2026 article on CMMS, intervention records and measurable reliability indicators, adapted for this technical guide. Read the original article.