Paper cup machine MTBF and MTTR can summarize reliability and maintainability, but the numbers are meaningful only when failure, operating time and repair boundaries are consistent. A better monthly value may reflect coding or production changes rather than equipment improvement.
This guide explains measurement discipline, not performance targets or warranty commitments. The site must define metrics appropriate to its assets and decisions. Always publish the calculation period, event count, exposure and exclusions with the reported KPI. Preserve the underlying events and revisions.
Diagnostic decision table
| Metric input | Definition needed | Distortion |
|---|---|---|
| Failure count | Included functional failures | Exclude inconvenient events |
| Operating time | Actual exposure basis | Use calendar time blindly |
| Repair start | Waiting versus active boundary | Mix logistics and wrench time |
| Repair end | Verified service restoration | Stop clock before quality approval |
Define the asset and failure event
State whether the measure applies to one machine, line, subsystem or component. Define failure as loss of required function under stated conditions. Separate planned maintenance, material shortage, quality hold and external utility loss unless the decision intentionally includes them.
Use the same event hierarchy across periods. Repeated resets of one unresolved fault may be one failure episode or several events depending on the rule; document the choice and preserve raw events.
Choose a defensible exposure denominator
Use actual operating hours, cycles or produced units when they represent exposure to failure. Record standby and planned shutdown separately. Comparing machines with different product mixes or speeds may require stratification rather than one pooled number.
Do not infer long-term reliability from a short period with no failures. Report observation time and event count with the result. Small samples create wide uncertainty even when the calculation is arithmetically correct.
Separate repair time components
Define detection, response, diagnosis, waiting for access, waiting for parts, active repair, testing and quality release. MTTR may refer to active repair or total restoration time; label it clearly. Preserve the components so different improvement actions can be chosen.
Stopping the clock when maintenance leaves the machine ignores failed restart tests. End restoration only at the defined functional state. Track temporary repair separately if full approved capability has not returned.
Interpret trends with supporting evidence
Plot failures and restoration components over time. Review failure modes, configuration, maintenance changes, product mix and data completeness. An MTBF increase accompanied by more unclassified microstops may be a data problem.
Use medians and distributions where a few long events dominate the average. Review critical failures independently from frequency. Metrics guide questions; they do not prove root cause or technician performance.
Use KPIs for specific decisions
Connect measures to actions such as spare readiness, diagnostic training, redesign, inspection or vendor support. Set a baseline and verify change using unchanged definitions. Record benefits and unintended effects on quality or safety.
Share HANNAI event definitions, raw timelines, failure modes and configurations when technical review is needed. A headline ratio without event evidence is insufficient for machine diagnosis or contractual conclusions.
Related equipment and next checks
Review the paper cup production line and the related guide to paper cup machine recurring alarm analysis. The Engineering Notes archive connects these checks with wider machine planning.
Frequently asked questions
Can calendar hours be used for MTBF?
Only when they correctly represent the defined exposure.
Does MTTR include waiting for parts?
State whether the metric covers active repair or total restoration.
Is a month with no failures proof of reliability?
Report exposure and uncertainty; short observation can mislead.
Can MTBF identify root cause?
No. Analyze failure modes and evidence separately.
Review the evidence with HANNAI
Send your machine model, component identification and the observations described in this guide. Include the drawing or material reference and any before-and-after measurements so the required next check can be identified.