Record Nonproductive Time Without Counting the Same Lost Minute Twice

A material shortage can block four sewing operators while a power failure blocks the whole group. Both events belong in the record. Adding both sets of minutes, however, may count some people twice during the same period.

That error changes the apparent size of the loss. It can also change which problem receives attention. A useful nonproductive time report must show who was blocked, when the blockage occurred and which work calendar applies.

This guide builds a complete garment factory nonproductive time calculation from a small event log. The figures are fictional. They show how to handle breaks, overlapping causes and changing crew sizes without hiding the original records. The method measures recorded blocked operator time. It does not set a wage rule, a standard allowance or an industry target.

Define the loss before collecting minutes

Start with the question the report must answer. A supervisor may need to know how much scheduled operator time was blocked by shortages or faults. Maintenance may need the time when a machine could not perform its required function. Planning may need the time when the line could not deliver output.

Those measures have different boundaries. One machine can stop while its operator performs approved work elsewhere. A line may keep producing from a buffer while an upstream station waits. Record the facts needed for each decision instead of treating every stop as the same kind of loss.

For the worked case, an operator is blocked when a recorded event prevents their assigned production work and no approved alternate work occurs. The measure includes only the agreed production window. Planned breaks are outside that window. These are explicit study choices, not universal definitions of nonproductive time.

Required checks, safe cleaning, training and other support work need their own treatment. Do not label all nonsewing activity as waste. Agree which activities belong in the production plan and which remain visible in a separate support-work record.

ILOSTAT's working-time definitions include related work, some downtime and certain rest periods within hours actually worked. Hours paid form a separate statistical concept. A production loss code cannot, by itself, decide pay or whether a period legally counts as work.

Keep elapsed time and operator time separate

Elapsed time follows the clock. Operator time adds the exposure of the people within the chosen boundary. A ten-minute event affecting six operators creates sixty operator-minutes if all six remain blocked throughout.

The crew size must stay constant for that multiplication to work. If two operators leave for alternate work halfway through, split the event at that change. Use the people actually affected in each segment.

Leading2Lean's downtime metric documentation illustrates the use of operator count times downtime for a labour loss measure. Apply the principle with a clear crew basis. A single average headcount can hide changes that matter within an event.

Keep these labels beside the totals:

Do not add the rows. They describe the same events through different units and rules. None automatically measures lost garments. That requires evidence about work flow, buffers, output and the task that limits production.

Equipment effectiveness is also a different measure. The Lean Enterprise Institute's OEE definition combines availability, performance and quality. A blocked operator-time share cannot replace that equipment calculation.

Build a log that preserves the event

Give each event a stable ID. Record the line, date, start time, end time, affected operator IDs, observed condition and person who checked the entry. Add the shift or production-calendar version used in the report.

Keep observation separate from explanation. “No approved bundle at station” describes a condition. “Poor planning” proposes a cause that needs evidence. The log should allow an initial reason to remain provisional while the team checks the material and release records.

Record changes in the affected group. If a helper moves, a replacement arrives or an operator starts alternate work, the event may continue but the labour exposure changes. A child record can hold the new interval and crew without creating a false second incident.

Use full date-time values. A stop from 23:55 to 00:10 lasts fifteen minutes across two dates. A time written as 08:25 means eight hours and twenty-five minutes, not 8.25 decimal hours. Store elapsed seconds or minutes for arithmetic and use clock times for display.

Give open events a visible status. At a live reporting cutoff, calculate only through that cutoff and mark the duration as provisional. When the event closes, retain the correction history. Do not silently replace an unknown end time with the end of the shift.

Set the calendar and crew exposure

The example covers six operators, P1 to P6, from 08:00 to 09:00. All six belong to the group for the full hour. A planned break runs from 08:30 to 08:40 and is excluded from this production measure.

Each operator therefore has fifty included minutes:

Included time per operator = 60 − 10 = 50 minutes.

Total included exposure = 6 × 50 = 300 operator-minutes.

The denominator comes from the roster and calendar, not from the stop log. A missing stop record must not make the scheduled exposure disappear. If people join or leave the group, calculate their included intervals separately and add them.

The same time rule must apply to losses. Leading2Lean's lost production time report distinguishes nonoverlapping downtime within active production schedules from a raw time report. Its daily and weekly report guidance also illustrates how a nonproduction period changes counted downtime. Those are software examples of a useful boundary check, not a required factory reporting convention.

Keep the break in the raw timeline. Exclude it only in the derived production measure. This preserves evidence for other reports that may use a different time basis.

Work through the four events

Assume the log is complete and every listed operator remains blocked during each included interval. No one performs alternate work in this first case.

Event B has fifteen raw minutes, but five lie within the planned break. Its counted interval ends at 08:30. Event C has twenty raw minutes. Its counted parts are 08:25–08:30 and 08:40–08:45, giving ten minutes in total.

The material and quality events do not cross the break. Their counted durations equal their raw durations.

Multiplying each counted duration by its affected crew gives:

Event A = 15 × 4 = 60 operator-minutes.

Event B = 10 × 6 = 60 operator-minutes.

Event C = 10 × 1 = 10 operator-minutes.

Event D = 8 × 2 = 16 operator-minutes.

The sum is 146 operator-minutes. It is a gross event total. It is not yet the unique blocked operator time because several event intervals overlap for the same people.

Merge intervals for each operator

For P1 and P2, the shortage runs from 08:10 to 08:25. The power stop overlaps it and continues to 08:30. Together they block each operator from 08:10 to 08:30, a single twenty-minute interval. The later quality hold adds eight minutes. Each person therefore has twenty-eight blocked minutes.

P3 and P4 share the same twenty-minute shortage and power interval. Neither has the later quality hold, so each has twenty blocked minutes.

P5 has the power stop from 08:20 to 08:30. The machine fault overlaps its last five minutes. The fault also adds five minutes after the break. P5 therefore has fifteen blocked minutes.

P6 has only the ten-minute counted power stop. The total is:

Unique blocked operator-minutes = 28 + 28 + 20 + 20 + 15 + 10 = 121.

The gross sum counted twenty minutes twice where the shortage overlapped power for four people. It also counted five minutes twice where the machine fault overlapped power for P5.

Overlap correction = 4 × 5 + 1 × 5 = 25 operator-minutes.

Unique total = 146 − 25 = 121 operator-minutes.

That subtraction works for this simple case. With three or more overlapping causes, pairwise subtraction can itself miscount shared time. Merging intervals for each operator is the safer general method.

Calculate the reported share

Blocked operator-time share = 121 ÷ 300 × 100 = 40.33%.

The other 179 operator-minutes have no recorded blockage under the example's complete-log assumption. They are not automatically earned standard minutes or proof of acceptable output. Those require separate production records.

Using the gross total would produce 146 ÷ 300 × 100 = 48.67%. The overstatement is 25 operator-minutes, or 8.33 percentage points of the included exposure. Round for display after calculating the unrounded values.

Calculate elapsed disruption only when needed

At least one operator is blocked from 08:10 to 08:30. After the break, P5 is blocked from 08:40 to 08:45, while P1 and P2 are blocked from 08:42 to 08:50. Those later intervals join to cover 08:40–08:50.

Any-operator blocked time is therefore twenty plus ten, or thirty elapsed minutes. It covers 60% of the fifty-minute included window.

All six operators are blocked together during the counted power stop from 08:20 to 08:30. That is ten minutes, or 20% of the included window. The blocked operator-time share remains 40.33%.

All three percentages are correct for their stated questions. They cannot be compared as if they used the same numerator. Label them in full and keep the underlying intervals available.

Even all-operator blocked time should not be renamed “lost output time” without checking the production boundary. A downstream group could still release completed goods. The event log describes the six people; the output record describes what the wider process delivered.

Allocate causes without adding the loss again

A cause chart needs one clear accounting rule. If each cause receives its whole event duration, the cause bars sum to 146 operator-minutes while the unique total is 121. That can be useful as gross involvement, but the chart must say so.

For an additive chart, assign each blocked operator-minute once. One possible rule gives priority to the cause that started first for that operator. It is an allocation choice, not proof of which cause mattered most.

Under this rule, the shortage retains its sixty operator-minutes. Power receives forty. P1–P4 add five power-only minutes each after the shortage ends, while P5 and P6 each contribute ten. The machine fault receives five minutes after the break, and the quality hold receives sixteen.

Attributed total = 60 + 40 + 5 + 16 = 121 operator-minutes.

A power-priority rule would instead assign forty to shortage and sixty to power. Machine fault and quality hold remain five and sixteen. The total still equals 121, but twenty operator-minutes move between the two leading causes.

Show the rule and its version beside the chart. Keep links to every overlapping event so investigators can see the full context. Do not delete the second cause merely because another cause owns that minute in the additive summary.

ASQ's Pareto guidance distinguishes measures such as frequency, time and cost. Use one measure per chart. ApparelDiary's Pareto guide for garment defects and rework minutes explains how the chosen measure changes the ranking. Here, the interval-allocation rule is another part of that definition.

Apply the method in a spreadsheet or system

Keep a raw event table and a separate calculation table. Preserve the original start and end times. Add calendar clipping and operator assignment in derived fields so someone can trace the result back to the event.

For each person, intersect the event with their included production intervals. An event may become two counted intervals when a break falls inside it. Discard only the portions outside the chosen exposure, keeping the raw record intact.

Sort the counted intervals by start time. Start a current interval with the first record. If the next interval begins before the current one ends, extend the end to the later end time. If it begins after the current interval ends, close the current interval and start a new one.

Intervals that touch at a boundary can be joined without adding time. Treat an interval as including its start and excluding its end. Thus 08:20–08:25 and 08:25–08:30 total ten minutes, with no extra minute at 08:25.

Add the merged durations for that person, then add across people. Check that no person's blocked total exceeds their included exposure. The group total must also stay within the group denominator.

A minute-by-minute grid can illustrate the method, but rounding each short event to a full minute changes the result. Retain the recorded resolution and round totals for display. If the source system only records whole minutes, state that limitation.

Reconcile the report before releasing it

Use three totals to check the worked case. First, the roster and calendar produce 300 operator-minutes. Second, the merged event intervals produce 121 blocked operator-minutes. Third, the remaining intervals contain 179 operator-minutes. The last two must add back to the first.

This check will not prove that every event was recorded. It will reveal arithmetic gaps or overlap in the way the included exposure was divided. Keep a separate check for source coverage, such as a missing device feed or a shift whose entries are still awaiting review.

Check cause allocation in the same way. The additive cause report must total 121, while a gross involvement report totals 146. Give each report its full label. A chart that switches between these bases from week to week can show a false change even when the events are identical.

Review duplicate entries before counting incidents. A supervisor and a mechanic may both record the same machine fault. Retain their observations but connect them to one shared incident identity where the evidence supports that link. Two independent faults close together may still be two incidents, even when their time intervals merge for labour accounting.

Keep incident frequency separate from duration. Splitting an event at a break creates two counted intervals, not two new failures. Splitting it at midnight for daily reporting also does not create a new incident. The original event ID ties the pieces together.

When several lines share a power stop, each line can retain its own affected crew. Consolidate operator exposure without assigning the same person to two lines at once. A plant incident count may be one, while six line records describe its local effects. Name the counting level so a manager can reproduce the total.

Before release, ask a second person to reconstruct one event from the raw times and roster. Then trace one cause total back to its operator intervals. This small review tests the links that a polished dashboard can otherwise conceal.

Make recording thresholds visible

A plant may choose to record only stops above a stated duration because manual entry takes time. If it does, the report measures those recorded stops. It must not imply that shorter interruptions were absent.

Keep the threshold stable when comparing periods. If a new sensor begins capturing short stops that the old paper log omitted, the measured total may rise even without a worse process. Review the change in coverage before drawing a production conclusion.

Document how the threshold works. Does it apply to the complete event or to each segment after a break? Is a short repeat linked to the earlier incident? These choices can change counts and minutes. Test them on a few real records before adding automatic rules.

The worked example uses complete records and no minimum stop duration. Its exact arithmetic should not be mistaken for evidence that a real factory log has the same coverage.

Check alternate work, missing data and corrections

Suppose P3 and P4 perform approved alternate work from 08:15 to 08:20 during the shortage. That removes five blocked minutes for each person under this article's definition. The shortage event still exists; its effect on labour has changed.

Unique blocked time becomes 121 − 10 = 111 operator-minutes. The share becomes 111 ÷ 300 × 100 = 37%. This is a revised scenario, not an extra deduction to apply to the original case without evidence.

Record the alternate task and its start and end. Otherwise a claimed redeployment can become an easy way to make losses disappear. Check that the person is not counted as available in two groups at once.

Treat missing records as missing. A failed tablet or unclosed event does not prove uninterrupted work. Report the affected exposure and investigate the gap. Do not present a precise loss percentage for a period whose denominator or event coverage cannot be established.

Use a short trial to compare operator, supervisor and machine records. A machine signal can confirm a stop but may not identify who was blocked or whether alternate work occurred. Resolve those differences before rolling the process out across many lines.

Discuss disputed causes with the people who did the work. HSE's consultation guidance describes two-way communication about health and safety. That is useful when reviewing stoppages linked to unsafe conditions. Its legal duties are specific to its jurisdiction; a loss target never authorizes unsafe work.

Compare periods and choose a response

A later observation may include five operators for fifty minutes each, giving 250 operator-minutes. Suppose it records ninety-six unique blocked operator-minutes. Its share is 96 ÷ 250 × 100 = 38.4%.

The count fell by twenty-five operator-minutes from the first case, or 20.66%. The rate fell from 40.33% to 38.4%, a reduction of 1.93 percentage points. Fewer people and a smaller denominator explain part of the count change. Neither comparison alone proves that a specific improvement caused the result.

Review the event pattern with style, staffing and material conditions. If shortages dominate, inspect release timing and bundle availability. If machine faults recur, link the events to maintenance records. ApparelDiary's MTBF and MTTR guide explains equipment failure and restoration measures without mixing them with labour effort.

Assign an action, an owner and a date for checking the effect. A reduction in recorded blocked time is useful only if the change supports safe, acceptable production. Check quality, output and any work moved to another group before calling it a gain.

Use ApparelDiary's factory KPI dashboard guide to connect the measure to a real decision. Show the period, crew basis, included exposure, unique blocked minutes and open-record status. Let users open the events behind a cause total.

Common questions

Should planned breaks count as nonproductive time?

That depends on the stated measure. This example excludes a planned break from both exposure and blocked production time. It preserves the raw event and makes no claim about pay, legal working hours or another report's definition.

Can one event affect several operators?

Yes. Keep one event identity and a record of the people affected over time. Expand it into operator intervals for the calculation. Count a person's minute only once when events overlap.

Does removing blocked time create the same amount of extra output?

Not necessarily. Output depends on the work sequence, constraints, available material and quality. Treat released operator time as a measured change in exposure, then verify the production effect with separate evidence.

Which cause should receive an overlapping minute?

Use a documented allocation rule for an additive report and retain every contributing event. Review the underlying causes through evidence. A priority rule makes the totals consistent; it does not settle responsibility.

References

• ILOSTAT: Wages and working-time concepts and definitions.

• Leading2Lean: Downtime-related production metrics.

• Leading2Lean: Lost production time report.

• Leading2Lean: Daily and weekly production summary reports.

• Lean Enterprise Institute: Overall equipment effectiveness.

• ASQ: Pareto chart.

• HSE: Consult your workers.