MTBF and MTTR: Calculate Both from Machine Event Records

Calculate observed MTBF as operating time divided by qualifying failures. Calculate MTTR from the restoration time for completed failure events. Define both clocks, the equipment boundary and exclusions before comparing machines or periods.

Two machines each lost eight hours last month. One stopped four times; the other stopped once. The maintenance lead needs different evidence for each: recurring fault causes on the first machine, and the long restoration sequence on the second.

Define the two clocks before calculating

IBM’s explanation of MTBF and MTTR separates operating time between failures from time needed to restore failed equipment. MTTR is used with several time boundaries in practice. Put your definition beside the value so a hands-on repair average is never compared with a stop-to-restoration average.

For this worksheet, a qualifying failure is an unplanned loss of the machine’s required function. Operating time excludes scheduled stops and failure downtime. Restoration time runs from that loss of function to the recorded return to the required function, including the intervening wait. Keep actual repair work as a separate duration. Apply your approved equipment definition consistently.

Compare two machines with the same total downtime

Illustrative scenario: both machines have 160 scheduled hours, eight hours of qualifying failure downtime and no other stops during the period. Each therefore has 152 operating hours. All failures started and were restored within the observation period.

Machine B has fewer recorded failures and a longer average restoration. Neither number establishes the remaining life of a component. One month and one failure are a thin basis for prediction. Retain the event count and observation hours whenever you present the averages.

Make the event log explain the result

Give each event a machine ID, failure ID, function lost, stop timestamp, notification time, maintenance arrival, repair start, repair finish, verification finish and restoration approval. The maintenance lead confirms the cause classification; production confirms when the required function returned. Link any work order and part issue used during the response.

If a fault recurs during testing before the machine is restored, preserve it inside the still-open event. If the machine was restored and subsequently fails again, record a new event. Decide this rule before review. Otherwise a team can make MTBF look better merely by combining several separate failures.

For the long event on Machine B, suppose the eight hours comprise three waiting for a spare, one for diagnosis, three for repair and one for verification. Those original, illustrative durations sum to eight. Buying a spare may address the wait; additional repair training would address a different part of the timeline. Keep concurrent activities from being added twice.

Handle incomplete periods openly

An event still open at month end has ongoing downtime but no completed restoration duration. Report it beside MTTR, including its current age. Do not omit that downtime from the operating-time calculation or assign a made-up finish time. For cross-period events, state which period owns the event count and completed restoration duration.

With zero qualifying failures, report zero failures in the observed operating hours. Division by zero does not produce a measured MTBF. Compare like machine duties, workloads and failure definitions; differences in product mix or planned maintenance belong in the review notes.

Accept the monthly report when every failure has an identity, the time categories reconcile to the reporting window, all completed events have approval evidence, and open events remain visible. Select an action against the actual repeated cause or delay and review the next comparable period.

Connect the clocks to production records

Our industrial IoT dashboards capture machine state and downtime context, while process entries connect work to production. Optiwise’s no-code relations and formulas can support your event register and calculations. A recorded stop still needs classification before it becomes a qualifying failure. Use the downtime tracking guide to establish the underlying event trail.

Maintenance calculation questions

Does MTBF include time waiting for a spare?

MTBF uses operating time, so time when the machine has failed is excluded. In this worksheet, the spare wait remains inside the event’s restoration duration and is shown separately for action.

Can an improving MTTR hide an unresolved breakdown?

Yes. An average based on completed events can fall while a long event remains open. Show open-event ages and total failure downtime beside the completed-event average.

Bring a month of machine events to an Optiwise demo to map the timestamps, classifications and review record.