Remap downgraded fabric digitally to avoid re-production

Submitted through the COP31 Sustainable Transformation Awards · SKD Türkiye (WBCSD Global Network Partner)

The company manufactures denim fabric at an integrated mill in Türkiye and employs between 251 and 1,000 people.

Re-production is the standard answer in textile manufacturing when fabric fails inspection, and it is the most resource-intensive answer available. Fabric classified as second quality on the quality control line was either scrapped or sold at low added value, and the resulting shortfall against the customer's order had to be closed by producing the same metreage again from raw material, with the full water, energy, chemical and labour consumption that entails.

The waste in that model is not the defect itself but the decision that follows it. A roll downgraded because of localised surface faults is largely sound material, yet the whole roll is treated as a loss, and the judgement about what to do with it rests on individual operator initiative rather than on analysis of where the faults actually sit.

The company set out to raise resource efficiency in production by removing that decision from human initiative and placing it on an algorithm. The initiative went live in 2020 and has run continuously since, and it is positioned as a twin transition project: the digital change and the environmental result are the same change, not two agendas run in parallel.

External pressure has since reinforced the case. Sustainability requirements such as the European Green Deal and the Digital Product Passport are pushing manufacturers towards digital solutions that deliver resource efficiency, which is the market the approach was built for.

Location of the initiative: Malatya, Türkiye

The approach converts fabric classified as a loss into fabric that re-enters the order, using an entirely domestically developed software solution that the company put into service in 2020.

The principle is a change of category rather than a change of process. Rolls that the quality control line classifies as second quality are analysed digitally at micro level, their surface faults are mapped, and the roll is remapped so that the sound areas are identified precisely rather than being written off with the defective ones.

The software then generates optimum cutting plans against that fault map, so that the usable material is recovered as first-quality product and returned to the production system. Value is created without any additional production: the metreage the customer ordered is delivered from material that already exists rather than from a new run.

Decision-making is removed from individual judgement. The traditional model relied on an operator deciding what to do with a downgraded roll; the software takes the decision to algorithmic accuracy on the basis of the fault map, which is what makes the result repeatable across shifts, fabric types and years.

The circular economy principle is therefore embedded in the production line itself rather than added at the end of it. What the industry treats as waste is treated as potential, and the recovery happens inside the mill's normal flow rather than in a separate recovery operation.

The core algorithm is not tied to a particular fabric or site. It works independently of fabric type and geography, and the same logic applies in any sector that generates a surface fault map and loses value to quality downgrades, which makes it a digital decision engine rather than a textile tool.

Figure 1: Quality optimisation screen for a second-quality roll: the fault map with the affected sections marked in red, the cutting instructions the software generates from it, and the recovered metreage and quality parameters of the resulting first-quality piece.

Fabric recovered from re-production year by year, 2020 to 2026, with the financial gain, water and carbon it represents. Total gain is calculated from the price difference between first and second quality.

Sustainability impact

Climate

The initiative targets Scope 2 and Scope 3 emissions, quantified using the GHG Protocol.

The Scope 2 effect is the purchased electricity that the avoided re-production would have consumed on the mill's own spinning, weaving, dyeing and finishing lines. The Scope 3 effect falls under Category 1: Purchased goods and services, covering the yarn, chemicals and other purchased inputs embedded in fabric that no longer has to be made a second time.

The base year is 2020, the year the system went live, and the impact data covers the six years of uninterrupted, verified production from 2020 to 2026. Over that period, preventing the re-production of 250,000 metres of fabric avoided approximately 1,914 tonnes of CO2 emissions.

The methodology values the recovered metreage by calculating the raw material, energy and labour equivalents of the fabric that was not produced, and applies the GHG Protocol together with textile sector water footprint standards to convert that into environmental terms. The underlying data is reportable through the company's enterprise resource planning system and its quality management reports.

Nature

Avoided re-production is also avoided water consumption, which matters in a sector where wet processing dominates the water footprint of a finished fabric.

Over the 2020 to 2026 period, preventing the re-production of the recovered metreage avoided approximately 7.9 million litres of water consumption. The figure is derived from textile sector water footprint standards applied to the fabric that was not produced.

Raw material extraction and processing are avoided on the same basis. Because the recovered metreage is delivered from material that already exists, the cotton, yarn and chemical inputs that a second production run would have required are never drawn, and the associated land and water use upstream is avoided with them.

The additional water consumption and carbon emissions from supplementary production are, in the company's terms, brought to zero for the volume recovered, since no additional production takes place at all.

Social

The change alters how quality is understood on the production floor. Fabric that the industry has traditionally treated as an error and a loss is shown to be recoverable through digital analysis, which shifts the mindset of everyone from suppliers to brands towards a zero-waste framing.

The quality controller role was strengthened rather than removed. The position is defined as the primary authority applying the dynamic cutting plans the system produces on the shop floor, so the operator moves from making a discretionary write-off decision to executing an analysed recovery plan.

Because the recovered metreage no longer has to be produced again, the pressure that re-production places on schedules — and the lost time it consumes — is removed from the operation.

Business impact

Benefits

The financial result is a net gain of USD 630,000 over six years, obtained from savings across four cost lines: raw material, energy, lost time and labour that re-production would have consumed.

The operational result is 250,000 metres of fabric returned to the economy rather than scrapped or discounted. Before the change, second-quality fabric was either written off or sold at low added value while the order shortfall was covered by a fresh production run; afterwards, the same material is recovered as first-quality product and value is created without additional production.

The gain is structural rather than one-off. The system has run without interruption for six years across different fabric constructions, which demonstrates stability rather than a favourable single period, and the results are verifiable through the enterprise resource planning system and quality management reports rather than resting on project estimates.

Removing the decision from individual initiative also removes its variability. Recovery no longer depends on which operator is on shift or how experienced they are, and the outcome is consistent enough to be planned against.

A further commercial option is open: the software can be converted into a cloud-based service offer and sold to other manufacturers, particularly in regions with dense textile clusters, which would turn an internal tool into a revenue line.

Costs

The main cost was the development of the software itself, which was created domestically and specifically for the company rather than purchased, together with the technical infrastructure and human resources that senior management allocated as a priority when the project was approved.

Integration cost is the second element. The system is embedded in the mill's core enterprise resource planning processes as a compulsory control mechanism, which means the change touched production, logistics and planning flows rather than sitting alongside them, and that integration work is not avoidable if the control is to hold.

Operating cost is continuous rather than one-off. Monitoring, development and audit of the system are a shared responsibility of the information technology department, the quality control management and the sustainability function, so three functions carry a standing workload.

The principal risk for anyone replicating the approach is data integration incompatibility between systems in different plants, which the company plans to manage through application programming interface based, flexible integration layers. That is an adoption cost for a second site rather than for the original one.

Costs are contained by developing the software in-house, by reusing the existing quality control inspection data rather than creating a new measurement step, and by the fact that the benefit is realised as avoided expenditure on raw material, energy and labour rather than requiring a new revenue stream to justify it. Converting the software into a cloud-based service offer is the identified route to recovering development cost beyond the original site.

The company puts figures on that profile. The core algorithm was designed in-house and the application was written in Python by the internal software team, and it runs on the company's existing local servers, so no capital expenditure on hardware and no cloud subscription was required. The technical development and SAP/ABAP integration effort came to TRY 280,000.

Annual operating expenditure is negligible: with existing infrastructure and no third-party software licences, the system needs only routine internal information technology monitoring and maintenance. Against that investment the company reports a payback period of under one month.

Indicative abatement cost

The initiative produces a net financial gain rather than a net cost, so the abatement cost is negative: USD 630,000 of net financial benefit accompanies approximately 1,914 tonnes of avoided CO2 emissions and approximately 7.9 million litres of avoided water consumption over the six-year period. A cost per tonne is not published, because the financial gain is not allocated between the emissions, water and operating cost effects.

Impact beyond sustainability and business

Co-benefits

The core algorithm works independently of fabric type and geography, so it can be integrated into textile production plants of different sizes. Beyond textiles, it is a digital decision engine applicable to any sector that builds a cutting plan from a surface fault map and loses value to quality downgrades.

The approach creates a mindset change across the value chain. Demonstrating that second-quality product can be recovered through digital analysis encourages everyone from suppliers to brands to think in zero-waste terms, which is a systemic effect rather than a saving at one mill.

Regulatory momentum works in the same direction: the European Green Deal and the Digital Product Passport push manufacturers towards digital solutions that deliver resource efficiency, so early adopters accumulate the data and the process maturity that later requirements will assume.

The next stage identified is conversion into a cloud-based service, which would spread the benefit across textile clusters and industrial cloud platforms rather than keeping it inside one company.

Potential side-effects

The compulsory control mechanism is deliberately rigid. Second-quality fabric is automatically blocked in the system and the block cannot be lifted until the remapping algorithm has been run, which guarantees compliance but also means a software fault or a data outage stops fabric moving to production, logistics or planning. The dependency should be planned for.

Removing individual initiative removes tacit judgement with it. Operator experience about which faults matter for which end use is no longer expressed as a decision, so that knowledge has to be captured in the algorithm's rules if it is not to be lost.

The recovery route depends on there being demand for the resulting cut lengths. A cutting plan that recovers sound sections produces different piece geometries from a standard roll, and the benefit is only realised where orders can absorb them.

Replication at other sites faces data integration incompatibility between plant systems, which the company identifies as the largest risk to wider deployment.

There is also a reporting caution: because the result is expressed as avoided production, the environmental benefit is a counterfactual rather than a measured reduction in the mill's own consumption, and it should be presented as such.

Typical business profile

The model suits textile manufacturers that inspect fabric on a quality control line, classify output by quality grade and currently cover the resulting shortfall by producing the same metreage again.

It is applicable at different plant sizes, because the core algorithm works independently of fabric type and geography, and it extends beyond textiles to manufacturers in other sectors that generate a surface fault map and build cutting plans from it.

Maturity requirements are moderate on equipment and high on data: the plant needs digital quality inspection data and an enterprise resource planning system into which a blocking control can be embedded, since the result depends on the check being unavoidable rather than advisory.

Delivery engages information technology, quality control management and sustainability functions, with senior management sponsorship to place the system inside the corporate digitalisation and sustainability roadmap.

Approach

1. Quantify what re-production actually costs before designing anything: Calculate the raw material, energy, lost time and labour consumed when a downgraded roll is replaced by a new run — four cost lines — so that the value of recovery is established in financial and environmental terms rather than assumed.

2. Digitise the fault map at the quality control line: Record surface faults at micro level and by position on the roll, because the remapping decision depends on knowing where the sound material sits, not merely that the roll failed inspection.

3. Build the remapping and optimisation algorithm to the plant's own conditions: Develop the software specifically for the production line rather than procuring a generic package, and keep the core algorithm independent of fabric type so that it remains usable as the product range changes.

4. Generate optimum cutting plans automatically: Have the software convert the fault map into cutting plans that recover the sound sections as first-quality product, so that the output of the analysis is an executable instruction rather than a report.

5. Make the check compulsory inside the enterprise resource planning system: Block second-quality rolls automatically and prevent the block from being lifted until the remapping algorithm has been run, so that a roll cannot move to production, logistics or planning while the recovery step is outstanding.

6. Remove the decision from individual initiative: Position the system as a systemic gatekeeper rather than as a recommendation tool, because the 100 per cent compliance the company reports comes from the barrier being technical rather than procedural.

7. Assign a named operational owner on the floor: Define the quality controller as the primary authority who applies the dynamic cutting plans the system produces, so that the algorithm's output has a specific role accountable for executing it.

8. Report the result through existing systems and recognised methodologies: Draw the operating data from the enterprise resource planning system and quality management reports, and convert avoided production into environmental terms using the GHG Protocol for Scope 2 and Scope 3 and textile sector water footprint standards, tracking five headline indicators — recovered area, avoided CO2e, financial saving, and the water and electricity not consumed.

Stakeholders involved

• Project leads: Senior management and the board supported the project from the initial idea through to go-live and included it in the company's strategic digitalisation and sustainability roadmap. Corporate ownership was demonstrated by prioritising the allocation of budget, technical infrastructure and human resources needed to implement it, rather than by endorsement alone.

• Company functions: Monitoring, development and audit of the system are the joint responsibility of three functions: the information technology department, the quality control management and the sustainability department. The quality control department also worked with the project team on the conceptual design and through the testing phase, and that collaboration continues as the system develops. The quality controller role is positioned as the primary authority applying the dynamic cutting plans on the shop floor.

• Main providers: No external solution provider was engaged. The software is an entirely domestic, in-house development, which is why the core algorithm could be built around the plant's own quality classification logic rather than around a vendor's data model.

• Other: Wider deployment is planned through partnerships in regions with dense textile clusters and through industrial cloud platforms, under a service-based model. Customers and brands benefit indirectly, since orders are fulfilled from recovered material rather than from additional production.

Key parameters to consider

The system went live in 2020, which is also the base year for impact measurement, and the reported data covers six years of uninterrupted verified production to 2026. The initiative is therefore a proven industrial application rather than a short-term pilot.

Cumulative results over that period: 250,000 metres of fabric recovered from re-production, USD 630,000 of net financial benefit, approximately 1,914 tonnes of CO2 avoided and approximately 7.9 million litres of water not consumed.

Five headline indicators are tracked: the recovered area successfully converted from second to first quality, the avoided carbon emissions in tonnes CO2e under Scope 2 and Scope 3, the financial saving in USD from raw material, labour and energy optimisation, and the litres of water and kilowatt hours of electricity not consumed in the prevented production.

Data is reportable through the enterprise resource planning system and quality management reports, and the environmental conversion follows the GHG Protocol and textile sector water footprint standards.

The core algorithm is independent of fabric type and geography. The principal technical constraint on replication is data integration compatibility between plant systems.

The emissions, water and financial results are calculated with the company's internal tracking systems and methodologies, and have not been subject to external assurance or independent third-party verification.

Implementation and operations tips

Make the control compulsory, not advisory. The company attributes 100 per cent process compliance to the fact that a second-quality roll is technically blocked until the algorithm has been run, and a recommendation-based system would not have produced a result that held for six years.

Count what was not produced. The methodology values the recovered metreage by calculating the raw material, energy and labour the avoided production would have consumed, which is what turns an operational improvement into a reportable environmental result.

Use the systems already in place for evidence. Drawing the data from the enterprise resource planning system and quality management reports meant the six-year record was verifiable without building a separate measurement infrastructure.

Keep the algorithm independent of the product. Building the core logic so that it does not depend on fabric type is what allowed the same system to work across different constructions and is the precondition for taking it to other plants or sectors.

Plan the integration layer before replicating. Incompatibility between plant systems is the identified risk for deployment elsewhere, and an application programming interface based, flexible integration approach should be designed in from the start rather than retrofitted.

Frame it as one transition, not two. Presenting the work as a combined digital and green change kept the digitalisation budget and the sustainability objective attached to the same project, which is what secured senior management sponsorship.