Value Stream Mapping for Garment Orders: Separate Touch Time from Waiting

A garment order can spend two days between cutting and sewing even when both departments meet their hourly targets. A faster cutting cycle will not explain where those two days went. The delay might sit in a full rack, a missing release record or a plan that sends work forward before sewing can accept it.

Value stream mapping brings those links into view. The Lean Enterprise Institute describes a value stream map as a view of both material and information flow. It begins with the current conditions and develops a future state that the team intends to create.

For a garment factory, the useful result is a shared view of how one product family moves through the work. The map needs more than boxes and arrows. It must explain the time figures and show who tells each stage to act. It also needs evidence for each proposed change.

Choose a family and a clear boundary

Start with products that follow a similar route through common equipment. The Lean Enterprise Institute's product-family definition uses those shared processing steps and equipment, viewed from the downstream customer. A broad label such as “woven garments” may hide several very different routes.

For example, unwashed shirts and washed trousers may share cutting resources but differ greatly after sewing. One map could become hard to read if it tries to combine both. A small family with a common final route gives the first study a clearer purpose.

Write the entry and exit events as actions that someone can see and record. “Material released to cutting” is more precise than “start.” “Packing complete and the batch released to the finished-goods area” is more precise than “finished.” Name the records that establish each event.

A map with those boundaries leaves out fabric purchasing and customer delivery. It can show a delay within the factory. It cannot claim to show the full time from order to delivery. If the customer problem lies outside the boundary, extend the study or link it to a separate map.

Give the map an owner who can work across departments. NIST's value stream mapping guidance stresses cross-functional involvement. In this application, cutting, sewing, finishing, planning and quality each know part of the journey. The person who handles bundles may explain a transfer that no planning report records.

Agree what each time measure describes

The same word can hide different measurements. “Sewing time” might mean the standard labour content of one garment, the interval between two line outputs, or the hours needed to complete a batch. Those figures serve different purposes.

The Lean Enterprise Institute's cycle-time glossary distinguishes operator cycles, processing time and production lead time. It also separates value-creating work from other activity. These distinctions matter before the team fills in a data box.

Use an explicit description beside each number:

Here, “stage occupancy span” means the time the batch spent within that stage. It does not mean cycle time. A sewing batch can be on the line for four hours. Within that span, each garment may spend much of its time waiting between tasks.

Likewise, ten minutes with two people working at the same time can represent twenty labour-minutes but only ten elapsed minutes. Labour content and calendar time measure different things, so they cannot simply be added. Write the unit and what it follows beside each value.

Walk the route and test the records

The EPA's Lean and Environment Toolkit emphasizes direct observation and discussion with the people doing the work. A route sheet shows the intended sequence. The mapping walk checks whether that route matches the actual flow.

Follow a real batch with a clear identity through the chosen boundary. Note when it arrives, starts and finishes each step. Record its moves and releases too. Ask where it waits and what prevents the next action. Include returns to earlier stages rather than drawing only the normal forward route.

An electronic timestamp needs a definition too. A scan entered at the end of a shift may show when someone entered the data, not when the batch arrived. Check several records against the work. If the two events differ, keep both. Do not quietly treat the later scan as the movement time.

Mark estimates as estimates. When a move has no reliable timestamp, the first map can show that evidence gap. A made-up time produces a neat total at the cost of a false baseline. Short observation periods can then supply the missing detail.

One observed order gives a case, not the usual factory lead time. Repeat the walk across the shifts, sizes and order types that matter. Record why any unusual route occurred. A repaired lot and a first-pass lot may need separate views even when they share the same style code.

Build a timeline that adds up

The following example is fictional. It follows one batch of 120 simple garments through a common route. The batch keeps its identity and moves as a whole between the stages shown. Tasks within sewing can overlap. The batch-level stages in this table do not overlap.

The clock starts when its fabric and instructions are released to cutting at 08:00 on Monday. It stops at final batch release after packing, at 16:00 on Wednesday. This example has no holidays or clock changes.

The sum is 3,360 minutes, or 56 elapsed hours. It also agrees with the difference between the entry and exit timestamps. That second check catches missing intervals, duplicated waits and a wrong date.

The three stage spans total 480 minutes, or eight hours. The other intervals total 2,880 minutes, or forty-eight hours. No stage span overlaps another interval in this case. Their sum therefore counts each minute once.

The factory's normal shift window is 08:00 to 16:00. Two overnight closures account for thirty-two of the forty-eight hours outside the stage spans. The other sixteen hours fall within shift windows; these windows may still contain breaks. The map should not label all forty-eight hours as available production time lost through poor performance.

Separate movement, waiting and touch time

Closer observation shows that the first queue-and-transfer interval includes twenty minutes of physical movement. The second includes fifteen. Together, these intervals span forty hours. They contain thirty-five minutes of movement and 2,365 minutes of waiting.

Including the initial and final waits gives 2,845 waiting minutes across the mapped journey. Add thirty-five movement minutes and 480 stage minutes to recover the same 3,360-minute total. This detail helps the team choose a relevant action. A shorter trolley route could reduce movement. An earlier release decision would address a different delay.

The stage total still does not tell us how much work happened on one garment. Suppose a separate study follows one marked garment in this batch. It is worked on for two minutes in cutting and eighteen in sewing. Finishing, checking and packing add five minutes. These activity periods do not overlap.

Its recorded activity time is twenty-five minutes. The garment shares the batch's entry and release events. Divide twenty-five by 3,360: activity takes about 0.74% of its elapsed journey. This share describes one observed garment; it is not a factory benchmark.

It is also not automatically a value-added percentage. The five-minute final activity includes checking. The team has not yet decided which work elements add value for the customer. Necessary checks can consume time without transforming the product. Calling them waste does not give permission to remove a required control.

The eight-hour stage total represents about 14.3% of the batch's elapsed time. Calling that figure “touch time” would overstate what was observed. Both percentages can be calculated correctly while only one label describes the underlying data.

Keep stock coverage separate from a traced journey

A map often records the quantity waiting between processes. That count provides another view of flow, but it does not replace timestamps.

Suppose a separate stock check finds 2,400 garments of the selected family waiting before finishing. At a stated demand of 600 garments per operating day, that quantity represents four operating days of demand coverage. It does not prove that each garment has already waited four days or will leave four days later.

Age, priority, product mix and available capacity can differ within the stock. A snapshot also differs from an average taken over time. If the queue has been growing, today's count divided by yesterday's output may hide that change. It would not describe a stable flow.

Minitab Engage's mapping documentation allows inventory-time calculations to use different bases, including takt time or the next process's cycle time. That choice changes what the displayed time means. A number from software still needs a clear basis. The user must know what it can show.

For the 120-garment example, the twenty-hour waits come from observed timestamps. Do not add another inventory-derived delay for the same queues. The stock calculation may sit beside the timeline to help explain flow. Adding it to the timeline would count the same waiting twice.

ApparelDiary's guide to queuing theory and factory buffers develops the wider flow relationship. Use that analysis when the required flow and averaging assumptions are supported, rather than treating every single stock count as a measured lead time.

Preserve the quantity when the route branches

The worked timeline keeps all 120 garments together. Real orders often split. Some sizes may move ahead while others await a missing component. A single arrow can conceal that difference unless it carries both an identity and a quantity.

Suppose 116 garments pass the final check while four are held. Those quantities still add to 120. They do not show that the whole batch is released. If the four remain unresolved at 16:00, the example's complete-batch exit event has not occurred. The lead-time record must stay open.

An authorized partial release would create a different event: release of 116 garments. It could be useful to measure both time to first release and time to complete release. The report must name each measure. Substituting the earlier time without changing its label would make delivery performance appear better than it was.

Returns require the same care. If the four garments go back to sewing, retain their link to the original batch and record the return path. Their repair activity may occur while the other garments wait. Adding both elapsed periods as though they were consecutive would overstate the batch journey.

For a branched route, put the dated events in order. Show what must be complete before the branches join the next step. Summing every branch is not a valid general rule. The latest required completion may control the next start, while some activities happen in parallel. A small event table can explain that relationship more clearly than a crowded arrow.

Put demand and capacity on the same basis

A future map needs a demand pace. That pace differs from the time spent by one garment. The Lean Enterprise Institute defines takt time as available production time divided by customer demand.

For a separate daily capacity check, assume 420 net available production minutes and demand for 600 garments. Takt time is 0.7 minutes, or forty-two seconds per garment. The 420 minutes already leave out the stated breaks and other nonproduction periods. Do not subtract those periods again.

The system must supply the required quantity at that average pace over the stated period. It does not require one operator to perform all eighteen minutes of the earlier sewing activity in forty-two seconds. Several operations and workers can contribute to the finished output.

Compare the required pace with observed accepted output at a clearly named point. Record the product mix and available staffing. A short, smooth run may give a better output interval than the whole shift. Stops, style changes and rejected work affect the shift result.

Where several families share equipment, their time commitments must fit together. Giving each family the full available shift would promise the same capacity more than once. The map can show that conflict. A detailed resource plan may be needed to resolve it.

Show the instructions that make material move

The material path explains where the batch goes. The information path explains why it goes there at that time. Add who releases cutting, who selects sewing's next batch, and who confirms that finished work may leave the area.

For each instruction, show who sends it, who receives it and how often it changes. A printed daily list, a live system queue and an urgent phone call can all influence the sequence. If they disagree, the map should show how the current team resolves the conflict.

In the fictional case, all required finishing checks and the release record are complete at 10:00 on Wednesday. The record then waits for an authorized reviewer. This person's routine batch review is at 16:00. These are explicit assumptions for this case. If a required test result were still missing, the same six-hour interval would have a different cause.

There may also be more than one reason for a delay. A batch can wait for material during part of an interval and for approval during another part. Separate those periods when evidence allows. Assigning the full interval to both reasons would count the same delay twice.

Talk with the people using the instructions before choosing a remedy. A person may follow a late review routine because the form is often incomplete at the earlier time. Moving the meeting without fixing the evidence could just create another round of checking.

Design one testable future-state change

The current map shows a possible target: the final release wait. The proposed change is to make the completed record available at 10:00 and arrange an authorized review by 11:00. The same acceptance criteria and release authority remain in place.

This proposal assumes that the reviewer can make that time without delaying other required work. It also assumes that the completed record is accurate and that no new hold arises. Put these conditions beside the future-state time. They are part of the calculation.

If the proposal works, the final wait falls from six hours to one. The batch leaves at 11:00 on Wednesday, after fifty-one elapsed hours. The five-hour reduction is about 8.9% of the original fifty-six hours.

No production-stage time changes in this scenario. It does not claim a faster sewing method, lower labour cost or extra output. The map identifies a shorter release path for the same batch under the stated conditions.

The earlier forty hours of queue and transfer remain. They deserve study because they are long. That does not justify deleting them from the future map. Sewing and finishing may have other orders committed to the same resources. Moving this batch ahead could push another one back.

The useful next question is which queue can shrink without causing a worse delay elsewhere. That may require a change to release quantity, sequencing or setup work. It deserves its own evidence and trial rather than a target written into every queue box.

Retain quality and resource consequences

Less waiting can be valuable, but the future map should include the work needed to achieve it. Earlier review may require a change in a person's schedule or a better way to collect evidence. Those costs are part of the decision even when the physical route stays the same.

A proposed layout or process change may affect water, energy, material waste or worker conditions. Bring the relevant specialists into that review. EPA's toolkit supports adding environmental data to the map and considering those effects when planning improvements. A time saving should not hide a new waste stream or a problem with how work is performed.

Use a small number of relevant resource measures. For example, a revised pressing sequence might need an energy observation as well as a time study. A shorter elapsed span alone does not prove that the same order used less energy.

Keep measures within their own boundaries. Whole-factory electricity divided by one trial batch would include unrelated work. A better map can reveal that the required data are missing; it need not turn an unsuitable total into a persuasive-looking result.

Compare the trial with its promise

Before testing the earlier review, choose which batches qualify. State how their entry and exit events will be measured. Record whether the evidence was complete at 10:00, when the reviewer received it, and when release was authorized. Preserve any hold and its reason.

The trial should compare the full journey as well as the targeted release interval. A shorter last wait may coincide with a longer upstream queue. If only the improved box is measured, the team could miss a worse overall result.

Include quality outcomes and additional work. More frequent review might expose incomplete records sooner, which can be useful, but repeated correction effort should remain visible. A quicker release with unresolved defects would not meet the intended result.

Review several comparable batches before treating the new route as normal. Product mix, workload and staffing can change during the trial. Keep those differences in view when judging the results. A before-and-after improvement is evidence to investigate, not proof that the proposed action caused every minute saved.

ApparelDiary's practical guide to statistical process control explains how performance can be followed over time. The map supplies the process boundary and reasons to measure. Repeated observations show whether the result holds.

Leave a map that another team can use

A usable final map names its product family, boundaries, observation dates and units. Each major time figure can be traced to records or a clearly marked estimate. Current observations remain separate from future targets.

The owner should be able to explain the total without changing definitions halfway through. Batch occupancy is not garment touch time. Demand coverage is not a measured age. A required check is not removed merely because it appears in a non-value-creating interval.

After a successful trial, update the working method and observe the new flow again. Keep the earlier map so that the change remains visible. The next improvement then begins from the current evidence, including any limits that the first trial did not solve.

References

• Lean Enterprise Institute: Value Stream Mapping.

• Lean Enterprise Institute: Product Family.

• NIST Manufacturing Extension Partnership: Value Stream Mapping.

• Lean Enterprise Institute: Cycle Time and Related Terms.

• US EPA: Lean and Environment Toolkit, Chapter 3.

• Minitab Engage: Value Stream Map and Timeline Calculations.

• Lean Enterprise Institute: Takt Time.