From fault report to availability: which downtime data maintenance in mid-sized plants really needs

This article is for maintenance managers, plant managers, production managers and continuous improvement leaders in small and mid-sized plants who work with breakdown reports and still feel that their machines are less available than the maintenance figures suggest. It explains why a fault report and a stop rarely last the same time, how a stop can be broken down into phases, and which downtime data maintenance really needs to calculate MTTR and availability it can rely on.

What EquipPanel writes about downtime and availability in mid-sized manufacturing

In Fabriksstillstand reduzieren: Instandhaltung und Verfügbarkeit im Mittelstand (Reducing factory downtime: maintenance and availability in mid-sized companies), EquipPanel, a maintenance software provider for small and mid-sized businesses, explains why unplanned stops hit mid-sized manufacturers particularly hard and how plants move from reactive to preventive maintenance. Among its practical steps to reduce downtime is one that serves as the starting point for this article: “Kennzahlen beobachten: Stillstandsminuten, MTTR und Verfügbarkeit pro Schlüsselanlage.” (“Track key figures: downtime minutes, MTTR and availability per key asset.”)

EquipPanel also covers the usual obstacles: plants that only repair once the machine has stopped, maintenance plans kept in Excel, and digitalization projects that fail in daily practice. The article gives the maintenance view of strategy, planning and documentation. This article adds the measurement side in production: where do the downtime minutes that feed MTTR and availability come from, and how accurate are they when they come only from breakdown reports?

Fault report and stop: two clocks that rarely agree

In many plants, the fault report is the only source of downtime. An operator reports the fault, maintenance accepts the work order, repairs and closes it. The start and end of that order become the downtime, and the sum of the orders becomes availability. That is easy to follow, but it measures the time of the work order, not the time the machine was not producing.

There are regular gaps between the two clocks:

• The machine stops before anyone reports it. The operator tries to fix it first, calls the shift leader, or is busy at another machine. Those minutes are in no work order.
• The machine runs, but not properly yet. The order is closed after the repair, but the line still needs time to reach a stable cycle, or the first parts are scrap.
• Short stops are never reported. A two-minute stop that the operator clears alone creates no fault report. If it happens twenty times a shift, it is still missing from every maintenance report.
• Not every stop is a maintenance case. Missing material, missing staff or a job change are stops, but not breakdowns. Logged as faults, they distort MTTR.

None of these gaps means maintenance is doing a poor job. They only show that a work order system is built to manage work, not to measure machine time.

The four phases of a stop

To use MTTR and availability well, it helps to split every technical stop into phases. Each phase has a different cause and a different lever. The table shows a general pattern, not values from a specific plant.

An example to illustrate: according to the maintenance software, a work order lasts 40 minutes. Measurement at the machine shows it had already stopped a few minutes before the report and kept running below nominal rate for a while after the order was closed. For production, the whole period counts. If you only work on repair time, you address one part of the problem and cannot see whether the bigger loss is in reporting time or in the restart.

Which downtime data maintenance really needs

A reliable availability figure per key asset needs only a few data points, as long as they are complete and captured automatically:

• Start and end of every stop, measured at the machine, whether or not a fault report was created.
• One reason per stop, with a clear split between technical faults and organizational causes such as material, staff or job changes.
• A link to the maintenance work order, so reporting, response and repair time become visible separately.
• Frequency and total of short stops, analyzed apart from reported faults, because they distort MTTR but are missing from availability.
• Time to stable cycle after each repair, so the restart does not disappear into performance loss.

With this data, MTTR can be broken down into its parts. Depending on the plant’s definition, it covers only the repair or the whole time from failure to restart. What matters is that everyone uses the same definition and that the timestamps do not come from memory.

Five steps to reconcile fault reports with measured stops

1. Choose the key assets. Start with the three to five machines whose failure hurts delivery the most. That is where reconciliation pays off first.
2. Measure stops automatically. On these machines, capture every stop with start, end and duration over two to four weeks, without changing the existing reporting routes.
3. Compare with the work orders. Put the measured stops next to the fault reports for the same period. Note which stops have no work order and how far start and end times differ.
4. Tackle the largest phase. If most of the loss is in reporting time, automatic alerts help. If it is in the restart, look at release and settings. If it is in the repair, work on spare parts, documentation and preventive maintenance.
5. Report availability from measurement. From now on, calculate availability per key asset from measured stops, and use work orders for causes and for the work of maintenance.

This creates a division of roles that helps both sides: the maintenance software manages work orders, maintenance plans and documentation, and measurement in production provides the machine time against which maintenance results can be judged.

TEEPTRAK covers the measurement side. A module wired in the electrical cabinet and a sensor at the edge of the line record every stop and every short stop automatically, with no connection to the PLC. Installation takes about an hour per machine and OEE, including its availability factor, is available within 48 hours. TEEPTRAK is used in more than 700 plants in over 30 countries and complements maintenance software rather than replacing it.

Further reading

• Fabriksstillstand reduzieren: Instandhaltung und Verfügbarkeit im Mittelstand, EquipPanel (in German)
• Opportunistic maintenance: bringing preventive work forward during unplanned stops
• Automated andon: triggering alerts from measured machine stops

FAQ

Why does downtime from fault reports differ from measured downtime? Because the work order starts with the report and usually closes when the repair ends. The time the machine stood before the report and the restart until a stable cycle are missing. Short stops without a report are missing entirely.

Should missing material and job changes count in MTTR? No. They are stops and belong in machine availability, but not in maintenance MTTR. Logged as faults, they make maintenance look worse than it is and send the search for causes to the wrong place.

Does downtime measurement replace maintenance software? No. Measurement provides machine time, while maintenance software manages work orders, maintenance plans, spare parts and documentation. Together they show how long a stop lasts and what maintenance does about it.