Research libraryPrint on Demand, Sample on Screen
Chapter 14 of 14 · Part IV · Draft web edition
GFT recommends staged adoption with named evidence gates and a scorecard that stays inside the factory performance system.
The cloth is woven from this chapter's 5 exhibits: each weft is one exhibit's own data, normalised; each warp group is a section. The probe reads the exhibit's own value where it crosses. Procedural visualisation, not a measurement.
Read-out
Illustrative GFT adoption sequence extending the staged logic of the Cameroon textile programme. Positions 0 to 4 are symbolic phase boundaries, not calendar commitments. Phase 0 requires reconciled records and recoverable order states; Phase 1 requires physical and operational release; Phase 2 requires repeatability and retained-order economics; Phase 3 requires separate advanced-customisation validation.
Unit: illustrative evidence sequence, not elapsed time. Status: illustrative. Sources: S03, S18. Illustrative figures are teaching examples, not measured factory results.
Overture · 1 min
Expansion is justified only after a bounded apparel service has reproduced acceptable garments, fulfilled its original promises and covered the complete cost of retained orders. A successful demonstration, an attractive digital sample or a busy printer does not establish those conditions. GFT therefore recommends a sequence that begins with reversible learning, installs the measurement spine before pilot orders, releases only a physically qualified route and expands one controlled dimension at a time. Advanced customisation remains a separate validation problem. This chapter closes the proposed adoption loop with named evidence, accountable roles and explicit stop, outsource and repair options. It extends GFT's Cameroon textile programme as a decision framework, not as evidence of an installed on-demand factory or achieved commercial result.
14.1
The first capability to qualify is not a printer. It is the ability to define a garment, obtain a physical result against that definition and learn from the difference. A controlled technical pack, a supplier sample and a small outsourced run can test much of that ability before the business commits to an integrated production stack. This starting point preserves options. It allows the team to reject a material, revise a service boundary or change a fulfilment partner without also defending an equipment purchase. On-demand manufacturing concerns a structural change in the production system rather than a simple technological upgrade, so equipment ownership cannot substitute for that operating work.[S03]
Least irreversible does not mean cheapest or permanently outsourced. It means that the next commitment should buy information whose absence currently blocks a decision. If pattern quality is uncertain, pattern and physical sampling work has priority. If supplier print quality is uncertain, a matched strike-off and care test has priority. If buyers will not accept the proposed delivery promise, an order trial has priority. Purchasing machinery before answering those questions can make the unresolved work harder to see because the organisation now has an asset that appears to demand use.
GFT proposes an explicit capability ladder. The first rung controls specifications, material identity and approvals. The second qualifies suppliers through samples and short runs. The third operates a bounded service using the smallest practical combination of owned and outsourced stages. A later rung adds equipment only where observed volume, accepted output, support and retained-order economics justify it. Advanced parametric or image-derived customisation sits beyond that, with its own release protocol. A rung is not a calendar milestone. It is a scope that can be entered, retained, narrowed or left.
An outsourced comparator should remain visible even after equipment is acquired. It shows whether internal production improves the complete accepted-order outcome rather than merely shifting cost from a supplier invoice into labour, maintenance, idle capacity and recovery. Model-specific printer guidance, for example, includes routine ink-handling tasks that continue on a schedule, which demonstrates why intermittent use does not remove every operating obligation.[S31] That example applies only to the documented machine. The general lesson is to cost the actual chosen equipment's maintenance and support duties rather than assuming that low utilisation makes them disappear.
The least irreversible route also protects the quality of the counterfactual. If the pilot compares a new internal route with an unrealistically weak baseline, later expansion will rest on a manufactured advantage. The baseline should be the route that could genuinely deliver the same garment service, with its actual outsourcing, transport, approval and recovery work recorded. A decision to retain that baseline can be a successful pilot outcome when it prevents a poor commitment. The objective is a defensible service, not a predetermined purchase.
14.2
Related · chapter 109 paragraphs step it
Read-out
Illustrative GFT control design informed by the order-driven production concept and by internal programme principles G02 and G03. States are proposed, not implemented software or measured cycle times. A job-completed event does not prove buyer acceptance.
Unit: . Status: illustrative. Sources: S03. Illustrative figures are teaching examples, not measured factory results.
Read this exhibit in its chapter ↗A pilot cannot reconstruct reliable evidence after the orders have passed through an informal process. Before accepting them, GFT would freeze the product scope, approval states, metric definitions, material identifiers, cost fields and accountable roles. This is the illustrative Phase 0 extension of the programme's performance-management design: capture source events at their origin, preserve stable denominators and connect operating records to a regular management review. The measurement system is part of the service being tested. It is not a reporting layer to be added once adoption looks favourable.
The product-scope record names the garment family, construction route, material or blank, print process, size coverage, artwork limits, intended care and permitted customisation. It also states what is excluded. An exact repeat, a revised artwork and a new style should not enter one undifferentiated style count. Their development work and approval obligations differ. The record should assign each a stable identifier so that later evidence can show which version actually passed and which buyer promise it supported.
The order record follows commitment, approval, release, production, dispatch, acceptance, cancellation, refund and subsequent purchase as distinct events. A print job identifier cannot replace an order identifier, and a dispatch scan cannot replace buyer acceptance. The material record links fabric or blank lots, inks, pretreatment and other relevant consumables to the released specification. The cost record captures development labour, setup, failed work, production labour, energy, packaging, delivery, recovery and period capability in fields whose boundaries match the economic model.
Responsibility needs to be assigned at the event, not only at the dashboard. A technical role approves the specification. A production role releases and reconciles physical work. A quality role records holds and acceptance evidence. A fulfilment role closes delivery and return events. A commercial role confirms net retained sales and subsequent commitments. A management reviewer checks that the same definitions survive the weekly and monthly reviews. Small teams may combine roles, but the record should still show which accountable person performed each decision.
The illustrative Phase 0 is ready for pilot orders only when two practical demonstrations pass. First, a sample record must reconcile the approved design, pattern, material, process route, physical tests and decision history without relying on memory. Second, a trial order must survive an interruption test. The team should be able to stop and restart the record, identify what has already been produced, prevent duplicate release and recover the commercial and physical state. This does not require simulating every possible failure. It requires proving that the minimum control system works before real orders make omissions costly.
The management-control literature supports placing risk assessment and corrective action inside the operating system rather than treating them as a separate compliance exercise.[S18] For this service, that means every failed test, late order, unsupported release or unresolved record receives a cause, an action, an owner and a review state. The measurement spine should make those failures visible while they can still change the decision. A pilot that hides its exceptions has not established repeatability; it has established only that the summary was easier to produce than the evidence.
Exhibit 56 expresses the proposed sequence as a dependency map. The illustrative numeric positions are symbolic stages, not months, promised dates or resource-loaded scheduling. Each later phase begins only after the preceding evidence gate opens, and a failed gate can return the work to an earlier phase or end the route.
The roadmap deliberately leaves duration open. Evidence may accumulate at different rates for a stable repeat product and a difficult new material. It also prevents a common misunderstanding of phased adoption: passing time does not open a gate. Only the recorded result does. If the service changes its material, route or order promise, the reviewer must decide which earlier evidence remains valid and which phase the change has reopened.
14.3
The illustrative Phase 1 authorises a bounded service, not a general claim about digital production. Its launch candidate must name the exact garment family, material route, print process, size coverage, customisation limits, approval bundle and delivery promise that passed. Fit, colour, care, safety and order recovery are separate gates. A favourable result in one cannot compensate for an unresolved failure in another. The launch record should show both the passed conditions and the failed or untested conditions so that nobody mistakes a narrow release for a universal capability.
The illustrative Phase 1 authorises a bounded service
illustrative
Physical fitting remains necessary where the intended digital workflow cannot resolve every technical issue. The virtual-versus-physical fitting study supports that bounded role for simulation rather than automatic replacement of a real fitting.[S07] GFT would therefore specify which fit decisions the digital sample may inform, which wearer or form checks remain required and who interprets disagreement. The launch package should include the pattern revision, material state, avatar or body basis, poses, physical observations and approval decision. Changing any of those inputs may require a new comparison rather than reuse of the earlier pass.
Care evidence needs equal precision. ISO's public catalogue scope notes that industrial and hospital articles can require special laundering procedures that are more severe in some respects, so the intended use must shape the test route.[S10] The rubbing scope distinguishes dry and wet tests, which prevents one visual assessment from carrying both decisions.[S11] The domestic washing and drying scope states that a complete test consists of both procedures, making the drying route part of the evidence rather than an omitted detail.[S12] These catalogue entries define method scope. They do not provide a buyer's acceptance grade or prove that the proposed garment passed anything.
The launch dossier should consequently identify the laboratory or competent reviewer, exact procedure edition, specimens, conditions, observations and agreed acceptance criteria. It should distinguish a booked test from a completed test, and a completed test from a passed requirement. Where a public catalogue scope is all the team has accessed, the full protocol still needs to be agreed with the qualified testing party. GFT's recommendation is to keep those limitations beside the release signature rather than in a general annex that nobody reads when an order is waiting.
Safety and process recovery sit in the same launch decision. IFC textile guidance identifies heat consumption in drying and curing and addresses occupational risks across textile operations.[S17] The proposed service should therefore qualify its actual thermal, chemical and moving-equipment controls using the selected process documents and site conditions. It should also test the order-recovery procedure after an interruption: what material was consumed, which stage is safe to resume, whether a release retry could duplicate work and how the buyer promise will be reviewed. A route that produces a good sample but cannot recover its records and physical state is not ready for repeatable orders.
A launch gate can have four outcomes. A full pass releases only the stated scope. A conditional pass narrows that scope, for example by excluding an unqualified size or material, and records the condition in the offer. A failed but repairable gate returns to development with a named cause and retest requirement. An unacceptable physical or safety result stops the affected route regardless of apparent demand or contribution. Conditional release must never become a quiet permission to sell the excluded condition.
The launch reviewer should sign a gate sheet that identifies each evidence object, its source, its owner, the decision and the remaining limitation. The sheet should also list the changes that automatically reopen review: material or blank substitution, ink or pretreatment change, pattern revision beyond an approved parameter, new care requirement, altered print placement, different subcontractor or a changed delivery promise. This change control protects the meaning of the word proven. The service that expands must be the service that passed.
Interlude · the cloth half-woven
Woven: sections 14.0 to 14.3Still to weave: 14.4 to 14.7
14.4
Related · chapter 098 paragraphs step it
Read-out
Illustrative GFT sensitivity, not a demand or cost forecast. Full practical volume = 200 retained orders/month; pre-recovery contribution = 10,000 XAF/order; recovery = 35,000 XAF per remade order; fixed period cost = 600,000 XAF. Cell = 200 x utilisation x (10,000 - remake rate x 35,000) - 600,000, divided by 1,000. Sources support thermal, pretreatment and maintenance cost categories, not these values. Additional capacity consumed by remakes is excluded and requires separate measurement.
Unit: thousand XAF per illustrative month. Status: illustrative. Sources: S17, S30, S31. Illustrative figures are teaching examples, not measured factory results.
Read this exhibit in its chapter ↗The illustrative Phase 2 joins two different kinds of evidence. Operational repeatability asks whether the controlled process can reproduce accepted garments and reliable fulfilment without exceptional recovery. Commercial retention asks whether orders survive the defined cancellation, refund and return window and cover their complete costs. A later buyer commitment adds evidence of continuing usefulness, but it does not repair a negative contribution or a failed physical gate. Expansion requires the relevant measures to hold together for the qualified scope over a pre-agreed observation window.
Repeat purchase needs a denominator that cannot be confused with production output. The eligible population is the set of earlier buyers whose orders reached the defined retained state and whose subsequent-purchase observation window has closed. Returning-buyer incidence counts eligible buyers who place a new commitment, while repeat-order frequency can count their subsequent orders. A remake, replacement or correction belongs to the original order, not to either repeat measure. These definitions should be fixed before the cohort is reviewed and retained when later cohorts are compared.
Operational evidence should include new and exact-repeat work separately. An exact repeat against an unchanged package tests whether the released process can reproduce its result. A new style within the qualified family tests whether the control system transfers to bounded variation. A new material, construction or care requirement may fall outside the qualified family altogether. Counting all three as equivalent expansion evidence would hide where the process actually learned. The expansion register should identify which dimension changed and which evidence carried over.
Retained-order economics must reconcile more than print revenue. Net retained sales should absorb discounts, cancellations and refunds under the agreed definitions. Costs should include materials, consumables, direct labour, process energy, packaging, payment, delivery, setup and recovery, with period capability and added capacity shown in the period result. Accepted output and retained orders need separate denominators because an internally accepted garment can still be returned, while a refunded order can still have consumed material and labour. The economic review should include failed orders in its numerator rather than calculating margin only on the clean cases.
The worked example shows why a scale gate needs both an observed result and a sensitivity. A point estimate above the arithmetic coverage point may still be fragile when order mix, recovery or utilisation changes. The reviewer should ask which measured variation is plausible within the tested scope, then show whether the decision survives it. Where the sample is too small to support that variation, the correct scope is a longer or repeated bounded trial, not a confident extrapolation. Manual planning remains appropriate while order history is insufficient for more elaborate forecasting.
Expansion should proceed along one principal dimension at a time where practical. Adding styles tests product-development transfer. Adding sizes tests pattern, avatar and physical-fit coverage. Adding order volume tests practical capacity, support and working capital. Adding equipment changes fixed cost, process control and maintenance obligations. Combining all four makes it difficult to attribute the next failure or improvement. If commercial pressure requires several changes together, the review should treat the bundle as a new validation candidate with a fresh baseline.
Volume incentives must not erase the conditions that made the pilot acceptable. OECD guidance identifies inefficient production planning and poor purchasing practices among common factors associated with excessive working hours in garment supply chains.[S28] The scale review should therefore examine overtime, unplanned subcontracting, late specification changes and recovery hours beside delivery and contribution. An increase in accepted output achieved by transferring hidden strain to workers or suppliers is not evidence that the original service promise can be repeated responsibly.
14.5
The illustrative Phase 3 is not an automatic consequence of earlier sales. It is a separate technical and operational pilot for capabilities whose failure modes differ from a stable repeat service. Parametric patterns can make selected dimensions and construction relationships programmable, while image-to-pattern research explores the recovery of sewing structure from visual input. Neither result authorises automatic production. The earlier service evidence remains useful only where the material, pattern logic, construction, customer specification and physical approval route remain equivalent.
The illustrative Phase 3 is not an automatic consequence of earlier sales
illustrative
GarmentCode makes valid sewing patterns central to accurate digital simulation and eventual fabrication.[S32] That supports a controlled investigation of parametric components, not a presumption that every parameter combination is manufacturable. A supervised pilot should define which dimensions may vary, their permitted relationships, the construction details that must remain fixed and the patternmaker's review obligations. Each generated candidate needs dimensional readback, seam compatibility, grain and notch checks, and the physical evidence required by its risk. An out-of-range request should stop rather than force the software to produce an answer.
Image-derived patterns create an additional provenance problem. Sewformer research proposes a two-level Transformer network for sewing-pattern prediction from a single image.[S33] A source image does not, by itself, reveal every hidden seam, internal component, material property or construction intention. The research candidate therefore needs human reconstruction of missing design decisions before it can become a technical pack. The validation record should separate what the image showed, what the system inferred and what a patternmaker or buyer approved. Those are different evidence objects with different failure responsibility.
The illustrative Phase 3 test set should include held-out cases rather than only examples used to configure the model. It should cover the intended range of shapes, sizes, artwork and construction complexity, with failures retained in the denominator. Evaluation should compare the generated package with a competent conventional route on equivalent approved outcomes, including correction labour and residual physical iterations. A visually plausible result that requires extensive silent repair has not reduced development work. Its repair belongs in the measured workflow.
The illustrative Phase 3 test set should include held-out cases rather than only examples…
illustrative
Automatic release should remain prohibited in this proposed framework. A responsible technical role must reconcile the generated pattern, customer specification, material route and physical garment evidence before production. The system should also provide a usable manual fallback and preserve the prior approved version. If a model or software update changes outputs, the team needs a recorded revalidation decision rather than allowing the update to rewrite the production baseline unnoticed.
Commercial permission is equally specific. The service must know whether it may reproduce an uploaded image, retain body or garment measurements and reuse a generated design. Those questions require clear customer terms and qualified review of applicable obligations; they cannot be solved by a model confidence score. The illustrative Phase 3 expands only when its technical, physical, information and permission gates all pass. Otherwise the proven Phase 2 service should continue without the advanced option.
The illustrative Phase 3 expands only when its technical
illustrative
14.6
The scorecard in this illustrative roadmap should change in use, not in definition, as the service advances. During Phase 0 it exposes missing records and baseline uncertainty. During Phase 1 it shows whether the physical route and fulfilment controls hold. During Phase 2 it reveals repeatability, retained-order economics and capacity strain. During Phase 3 it makes correction work and new failure modes visible. Stable definitions allow those comparisons; different review questions prevent one summary metric from carrying every decision.
Exhibit 57 defines the illustrative measurement dictionary. Each measure needs a numerator, denominator, source record, accountable role and observation window. Baseline fields and acceptance thresholds remain deliberately unfilled because no local pilot observations have been established. A threshold should be agreed against buyer requirements, process variability and the decision at stake, not borrowed from a vendor demonstration or an unrelated study.
Read-out
| Physical iterations per approved style | All physical iterations, including failures, divided by approved styles | Technical development | Style cohort; compare complete development routes |
|---|---|---|---|
| Development hours per approved style | End-to-end development and correction hours divided by approved styles | Product development | Weekly and cohort; include digital and physical work |
| First physical approval incidence | Styles passing at first physical review divided by styles physically reviewed | Quality and pattern review | Cohort; diagnose by cause, never pursue as a quota |
| Retained-order contribution | Net retained sales less defined order costs, summed and shown per retained order | Commercial and operations | After return window; economic scale gate |
| Accepted units per staffed hour | Accepted garments divided by staffed hours on the complete route | Production | Daily and weekly; practical-capacity review |
| Delivery against original promise | Orders delivered by original promise divided by orders due | Fulfilment | Weekly; report approved amendments separately |
| Reprints and returns | Reprints divided by starts; returned orders divided by dispatched orders | Quality and fulfilment | Weekly and monthly; keep causes separate |
| Water per accepted garment | Allocated process water divided by accepted garments | Operations and sustainability | Defined period and process boundary; reconcile meters |
Illustrative GFT scorecard extending the programme's stable-definition, source-capture and review discipline. Baselines and thresholds must be agreed from local requirements and observations. Resource intensities are operational measures within stated boundaries, not product-footprint claims. No measured values or universal pass grades are supplied.
Unit: illustrative metric dictionary. Status: illustrative. Sources: S17, S18, S26, S28. Illustrative figures are teaching examples, not measured factory results.
Read-out
| Energy per accepted garment | Allocated process energy divided by accepted garments | Operations and maintenance | Include startup, idle and recovery states |
|---|---|---|---|
| Evidence reconciliation | Orders with complete required records divided by orders due for review | Performance-system owner | Every gate; incomplete evidence blocks expansion |
Illustrative GFT scorecard extending the programme's stable-definition, source-capture and review discipline. Baselines and thresholds must be agreed from local requirements and observations. Resource intensities are operational measures within stated boundaries, not product-footprint claims. No measured values or universal pass grades are supplied.
Unit: illustrative metric dictionary. Status: illustrative. Sources: S17, S18, S26, S28. Illustrative figures are teaching examples, not measured factory results.
Physical iterations and development hours belong together. A falling sample count can conceal growing digital correction, repeated buyer review or downstream remakes. First physical approval adds diagnostic value, but it should not become a target that pressures reviewers to accept weak work. The useful reading is by material, construction, style novelty and failure cause. A low incidence may identify a correctable preparation problem; an unusually high incidence with little recorded revision may identify missing evidence rather than excellent development.
Accepted units per staffed hour must use the complete route and a consistent staffing boundary. Printer runtime alone would omit preparation, curing, cutting, sewing, inspection and recovery. The staffing denominator should say whether indirect quality, maintenance and fulfilment hours are included, then keep that rule across compared periods. Delivery against the original promise should not be rewritten after a missed date. If the buyer agrees to an amendment, the original and amended outcomes can both be reported, but they remain different events.
Reprints and returns also need separate denominators and cause codes. A reprint concerns production recovery before the order closes; a return concerns the buyer or service outcome after dispatch. Both can consume material and capacity, and neither should disappear inside a general yield percentage. The review should connect their causes to a controlled change in specification, process, supplier or service terms. Resource measures need the same connection to physical work. IFC guidance supports attention to drying, curing and unsettled operating conditions, so startup, idle and recovery states should not be trimmed from the observed energy boundary.[S17]
Water and energy per accepted garment are operational intensities, not complete environmental conclusions. The historical Printscorpio case shows that batch and lifetime conditions can change a printing comparison and that its results belong to a particular system boundary.[S26] The scorecard should therefore publish the meter or allocation origin, process stages, accepted-output denominator and observation period beside every intensity. It should not turn those fields into a claim of national savings, low total impact or improved garment service life.
Review cadence should follow the decision. Source records are checked as events occur. Production, quality and recovery exceptions belong in short operational reviews. Cohort development, retained-order economics, resource reconciliation and working-hour risks belong in period and gate reviews. IFC's management-system guidance supports prioritised risk treatment, while OECD guidance keeps planning and purchasing consequences inside the responsible-operation review.[S18][S28] A dashboard supports that cadence only when every displayed number can be traced to its source rows.
The performance-system owner should issue a metric dictionary version whenever a definition changes. Historical results must retain the old definition or be recalculated transparently where source data permit. If a denominator cannot be reconstructed, the series break should be visible. GFT would rather show a missing baseline than populate it with a target borrowed from another operation. Evidence maturity is itself a pilot result: the team learns whether its proposed measures are collectible, intelligible and capable of changing a decision.
14.7
Qualified repeat service
Qualified repeat service: Evidence completeness 9, Consequence of incorrect expansion 5. Quadrant: Prepare bounded release · illustrative
The gate review should choose among proceed, narrow, repair and retest, outsource and stop.
Additional size coverage
Additional size coverage: Evidence completeness 6, Consequence of incorrect expansion 7. Quadrant: Prepare bounded release · illustrative
Integrated equipment purchase
Integrated equipment purchase: Evidence completeness 4, Consequence of incorrect expansion 9. Quadrant: Collect evidence · illustrative
Outsourced curing
Outsourced curing: Evidence completeness 7, Consequence of incorrect expansion 6. Quadrant: Prepare bounded release · illustrative
Automatic pattern release
Automatic pattern release: Evidence completeness 2, Consequence of incorrect expansion 10. Quadrant: Collect evidence · illustrative
The method behind it: Warning.
Illustrative GFT review positions, not measurements or probabilities. Horizontal position is completeness of the required evidence, where 0 is absent and 10 is complete. Vertical position is the consequence of an incorrect expansion decision, where 0 is low and 10 is high. Actual proceed, narrow, repair, outsource or stop choices require the named gate records.
Unit: illustrative review positions from 0 to 10. Status: illustrative. Sources: S07, S18, S28. Illustrative figures are teaching examples, not measured factory results.
14.7 · the bodyExhibit 59 above; its figures are illustrative where marked.
The gate review should choose among proceed, narrow, repair and retest, outsource and stop. These are different decisions with different evidence. A fit or care failure blocks the affected physical route even when its economics appear favourable.[S07] A support gap may justify outsourcing while internal capability is repaired. Negative retained-order contribution calls for price, product, process or service-boundary diagnosis. Delivery failure calls for capacity and purchasing review, not a larger sales target. Missing or irreconcilable evidence blocks every expansion claim. Risk prioritisation should direct corrective action rather than determine how much uncertainty to hide.[S18]
Exhibit 59 is an illustrative screening device, not a calculated risk score. It positions candidate decisions by the completeness of their evidence and the consequence of expanding without it. The detailed gate record remains decisive: no coordinate can override a failed physical, safety or working-condition requirement.
The final recommendation is a decision record tied to one qualified product family. It should state the approved scope, baseline, evidence period, gate results, contribution and cash boundaries, unresolved risks, named owner and next review trigger. It should also preserve the option to stop or outsource without treating that outcome as failure. Material, supplier, order-mix or service-promise changes reopen the relevant evidence rather than inheriting the previous approval automatically.
Exhibit 60 closes the illustrative operating loop. The next order uses a controlled revision, not an informal collection of lessons. Buyer acceptance and retained-order results return to the same specification and performance records that authorised production.
GFT would proceed only with the smallest service whose physical quality, fulfilment, retained-order economics and evidence reconciliation pass together. Expansion then follows measured repeatability, one bounded change at a time. If those conditions fail, the accountable decision is to narrow, repair, outsource or stop before commitment grows.
Check the record
Every knot is a source this chapter cites; select one to read its record, or filter the register by grade. Publication dates and inherited-source limits stay explicit. Read the paper's full evidence summary.
12 sources · 29 citation marks on this pageMost cited: S18, 4 times
Grade AGrade BGrade CKnot size: citations on this page
All 12 records shown
c14-S03 · Grade B
On-Demand Manufacturing Technology in Textile and Fashion Industry
The findings suggest that on-demand manufacturing represents a structural transformation rather than a simple technological upgrade.
2025-12-31. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c14-S07 · Grade B
Virtual Fit vs. Physical Fit: How Well Does 3D Simulation Represent the Physical Reality
It cannot fully replace a real physical fitting, since not all technical issues have been solved yet.
2020. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c14-S10 · Grade A
ISO 105-C06:2010: Colour fastness to domestic and commercial laundering
Industrial and hospital articles may be subjected to special laundering procedures which may be more severe in some aspects.
2010. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c14-S11 · Grade A
ISO 105-X12:2016: Colour fastness to rubbing
Two tests may be made, one with a dry rubbing cloth and one with a wet rubbing cloth.
2016. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c14-S12 · Grade A
ISO 6330:2021: Domestic washing and drying procedures for textile testing
A complete test consists of a washing and drying procedure.
2021. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c14-S17 · Grade B
Environmental, Health, and Safety Guidelines: Textile Manufacturing
Heat consumption is particularly significant in drying and curing operations and in activities involving wet treatments.
2007-04-30. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c14-S18 · Grade B
Environmental and Social Management System Implementation Handbook: Textiles and Apparel
You can use the Risk Assessment Form to prioritize which risks should be addressed first.
2014. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c14-S26 · Grade B
Environmental performance of future digital textile printing: Case study Printscorpio
When smaller printing batches are studied, digital printing causes less greenhouse gas emissions than screen printing.
2014. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c14-S28 · Grade B
OECD Due Diligence Guidance for Responsible Supply Chains in the Garment and Footwear Sector
Three of the most common factors are low wages, inefficient production planning and poor purchasing practices.
2018. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c14-S31 · Grade C
SureColor F2270 User Guide: Shaking the Ink Packs
Shake the White ink supply units once a day, and shake the color ink supply units once a month.
2026. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c14-S32 · Grade B
GarmentCode: Programming Parametric Sewing Patterns
Realistic representation of garments with valid sewing patterns is key to their accurate digital simulation and eventual fabrication.
2023-06-06. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c14-S33 · Grade B
Towards Garment Sewing Pattern Reconstruction from a Single Image
Then, we propose a two-level Transformer network called Sewformer, which significantly improves the sewing pattern prediction performance.
2023. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗S03 The Korean Fashion and Textile Research Journal. 2025-12-31.
S31 Epson America. 2026.
S18 International Finance Corporation. 2014.
S07 3DBODY.TECH conference proceedings. 2020.
S10 International Organization for Standardization. 2010.
S11 International Organization for Standardization. 2016.
S12 International Organization for Standardization. 2021.
S17 International Finance Corporation, World Bank Group. 2007-04-30.
S28 Organisation for Economic Co-operation and Development. 2018.
S32 arXiv, Korosteleva and Sorkine-Hornung. 2023-06-06.
S33 arXiv, Liu and co-authors. 2023.
S26 VTT Technical Research Centre of Finland. 2014.
Keep the thread · chapter 14 of 14
Move the shuttle along the rack to preview any chapter: its sections, exhibits and reading time. Draft web edition. This research is not investment, legal or tax advice; illustrative models are not measured factory results.
The paper's rack · 14 boltsBolt length: reading time
01Part I: Confirm the order before choosing the machine
GFT argues for a measured order-to-garment service that reduces speculative commitments while retaining physical proof of quality.
02Part I: Confirm the order before choosing the machine
Choose what the buyer commits to and where customisation occurs before quoting a minimum order or delivery date.
03Part I: Confirm the order before choosing the machine
Select a process against the substrate, print area, quality requirement and true batch cost.
04Part II: Make the digital sample physically accountable
A controlled colour workflow links artwork, fabric preparation, ink, cure and measured output.
05Part II: Make the digital sample physically accountable
A technical digital sample needs accountable pattern geometry, body measurements and construction intent.
06Part II: Make the digital sample physically accountable
Link the digital material to a physical specimen, control the simulation experiment and qualify the decisions that survive comparison.
07Part II: Make the digital sample physically accountable
Separate decisions that can be made on screen from those that require physical evidence.
08Part III: Prove the service and the economics
Connect approved artwork and patterns to material yield, cutting, sewing and accepted output.
09Part III: Prove the service and the economics
Compare physical sampling and digital sampling, then compare production routes under matched quality and service conditions.
10Part III: Prove the service and the economics
Make every design revision, production release and fulfilment event recoverable and auditable.
11Part III: Prove the service and the economics
Measure the whole order-to-use system and show which impacts move beyond the printing stage.
12Part IV: Buy, pilot and scale against evidence
Choose a supported production system and qualify the team that must operate and maintain it.
13Part IV: Buy, pilot and scale against evidence
Pre-register the comparisons and failure conditions before approving adoption.
14Part IV: Buy, pilot and scale against evidence
GFT recommends staged adoption with named evidence gates and a scorecard that stays inside the factory performance system.