Prepare for textile ecodesign without inventing its final requirements

The ESPR has been in force since 18 July 2024. It replaces the earlier Ecodesign Directive and establishes a framework for setting requirements on specific product groups. The Commission describes coverage of virtually all physical goods, subject to limited exemptions, and a wider range of possible requirements than energy efficiency alone. That framework is the legal starting point. It is not, by itself, an apparel test protocol or a finished list of passport fields.[S12]

The 2025 to 2030 working plan explains how product-specific work will proceed. Ecodesign requirements may address product performance, product information, or both. They will be set through delegated acts, either product by product or horizontally for groups of similar products. Generic examples in the plan include durability, availability of spare parts, minimum recycled content, key product features and environmental footprints. These examples explain the regulatory mechanism. They do not assign every example to clothing or establish a numerical clothing threshold.[S13]

Textiles and apparel are prioritised in that working plan, with garments and footwear expressly included in the first priorities. The plan identifies potential improvement in product lifetime, material efficiency, water use, waste, climate impacts and energy consumption. This justifies early managerial attention. It does not prove that a particular garment has those impacts, that a particular intervention will reduce them, or that every possible measure will appear in the eventual act.[S13]

A useful preparation register therefore distinguishes the requirement family from its unresolved specification. Durability is a possible performance family. The applicable use scenario, sample, conditioning, test method, unit, acceptance limit and conformity procedure still need the relevant adopted rule. Recycled content is a possible family. The eligible input, chain of custody, calculation denominator and verification method still need definition. Footprint information is a possible family. Its system boundary, data quality, allocation and comparison rules cannot be invented by the supplier.[S12][S13]

The same discipline applies to information. The Commission lists possible textile DPP content such as product identification and characteristics, fibre composition where relevant, use and repair information, end-of-life information, origin information and identification of relevant economic operators. The list is explicitly conditional on the final delegated act. A supplier can organise evidence around these families without declaring that every listed item will become mandatory for every garment.[S11]

Exhibit 34 converts that distinction into a workshop map. The horizontal score records whether the opened sources state a principle, identify a conditional family, or leave the exact requirement unresolved. The vertical score records GFT's recommended preparation response. Both are derived facilitation scores, not legal ratings, measurements of readiness or predictions about the final act.[S11][S12][S13]

The practical conclusion is not to wait for complete certainty. Product identity, specification control and traceable test evidence have operational uses before the legal mapping is finished. Conversely, a supplier should not hard-code a guessed threshold or publish a conformity statement merely because a requirement family appears in a policy document. Build the evidence, preserve its provenance, and leave the legal acceptance rule as a separately versioned object.

This also prevents a category error in procurement. A laboratory can test a declared property under a declared method. A software supplier can store and present fields. Neither activity alone establishes that the test or field is the one required by a future textile act. The purchase specification should say which proposition each service supports and which decision remains with the legal reviewer and the responsible economic operator.

The Commission's textile page gives Q4 2027 as the planned adoption of the ESPR delegated act for textiles. It immediately qualifies the timetable: implementation timelines may evolve as legislative and technical work progresses. It also places additional guidance, technical specifications and implementation measures after adoption. Planned adoption is therefore not a universal date by which every textile product must carry a compliant passport.[S11]

Exhibit 35 preserves the sequence without turning it into a fabricated deadline. The first event is the framework entering into force. The second is the working plan that programmes the product-specific work. The third is planned textile-act adoption. Later legal and technical milestones remain conditional on adopted instruments rather than being inferred from the quarter shown in an indicative plan.[S11][S12][S13]

A legal-change register should give each milestone its own row. Record the instrument, product scope, status, source date, adoption date where established, publication status, entry into force, transition, application and responsible reviewer. Keep consultation material and technical guidance in separate status classes. A preparatory study can inform design; it should not silently become the source of a binding product limit. The working plan itself is policy programming, not the textile delegated act.[S13][S11]

Investment approval needs the same separation. A supplier can authorise record cleaning, specification control and evidence retrieval because those activities answer present operating needs. Expenditure tied only to an unadopted field or interface should carry a change allowance and a stop point. The reviewer should be able to explain which part of the work survives a different final schema and which part would need rework. That is a more defensible contingency than a precise budget built around an assumed legal deadline.

The register also needs reopening triggers. A new draft or adopted act, a technical specification, a changed product family, a different sales route and a buyer schema revision can each alter the assessment. Assign the monitoring task to a role and retain the text used for each decision. When the timetable moves, update the plan rather than preserving an obsolete countdown in sales material or a factory dashboard.

The Commission states that the textile DPP will be accessible at online points of sale and through a data carrier on the product, such as a QR code. It also distinguishes access to ecodesign information from authority access to conformity-assessment information required under applicable legislation. The carrier is therefore an entry mechanism with different information needs and audiences. Its precise form, granularity, access rights and technical specifications remain subject to the final requirements.[S11]

A working QR demonstration proves a narrow technical proposition: a code resolved to a page during a test. It does not prove that the code identifies the correct product population, that the displayed evidence is complete, that the user is entitled to the content, or that the page will persist for the required period. Nor does placing the same record on a blockchain answer those questions. Evidence quality and legal applicability remain separate from storage novelty.

Exhibit 37 sets out a proposed demonstration path. The failure branches matter as much as the successful page view. An unresolved identifier should stop publication. An unauthorised request should receive only the permitted view. A changed source record should create a reviewable update rather than silently replacing what a buyer or authority previously saw.[S11]

The test protocol should use known cases and record their expected results. Include a valid product, a superseded revision, an unknown identifier, a user without authority access and a source document under correction. Check whether the page states its product scope and evidence date. Test the export received by a buyer as well as the screen viewed by a consumer. A visually clear page can still conceal a broken evidence link.

Persistence needs an owner and a recovery path. Record who controls the domain or resolver, how identifiers are retained across a system migration, and what happens when a product is discontinued or a supplier changes. Do not invent a statutory retention period before the relevant act specifies one. The present objective is to avoid a demonstration whose links fail as soon as the vendor, product name or website structure changes.