JRC 145830 vs CIRPASS D2.1: the technical divergences textile brands must resolve before the delegated act

Two parallel technical references

The European textile industry faces a challenge of technical synchronisation. Regulation (EU) 2024/1781 on ecodesign for sustainable products (ESPR) establishes the unavoidable legal framework. However, the implementation of the Digital Product Passport (DPP) depends on specific delegated acts. The ESPR Working Plan 2025-2030, published in document COM(2025) 187 final, sets the 2027 horizon for the adoption of the textile delegated act. While the legislator defines the exact requirements, the engineering and compliance teams must build the data infrastructure based on preparatory documents that show critical divergences.

At the centre of this institutional gap operate two documentary blocks. On the one hand, the CIRPASS project delivered the D2.1 "Legal Mapping" report in July 2023 and the D3.2 "System Architecture v1.9" in May 2024. These texts define a conceptual architecture based on early pilots. On the other hand, the Joint Research Centre (JRC) of the European Commission published in March 2026 the report JRC 145830 "Methodology for defining data requirements for the Digital Product Passport under the ESPR framework" (DOI 10.2760/4511279). There is a difference of approximately 3 years between the start of the CIRPASS mapping and the final JRC methodology.

Both references coexist because they respond to different mandates. CIRPASS explored the initial technical feasibility. JRC 145830 directly instructs the legislators on how to draft the data specifications in the delegated acts. This duality obliges economic operators to understand which parts of the CIRPASS architecture remain valid and which components require immediate refactoring to pass the JRC's methodological audit.

Architectural agreements

The base topology of the system admits no debate. Both the CIRPASS approach and the JRC methodology conceive the DPP as a decentralised knowledge graph. The infrastructure must be anchored in a unique product identifier (Product UID) that resolves through persistent URIs.

The data model requires the use of Linked Data. The W3C Verifiable Credentials Data Model 2.0 (W3C VC 2.0) operates as the mandatory cryptographic security layer. A compliant document must be served under the media types "application/vc" or "application/vp", guaranteeing that each assertion of the passport includes a data integrity proof. Verifiable credentials prevent manipulation through selective disclosure schemes and the strict use of the @context property to map terms to semantic dictionaries (JSON-LD 1.1).

The CIRPASS D3.2 document establishes the use of Decentralised DPP Data Repositories (DDR). The responsible economic operator does not send a static file to a central server. It must expose a REST-API interface that can translate requests into SPARQL queries to interact with RDF databases. JRC 145830 assumes this same interoperable infrastructure. The agreement is absolute in the transport layer and in the applied cryptography. Textile brands can implement HTTP resolvers based on RFC 3986 or use Decentralised Identifiers (DIDs) without risk of obsolescence in the architectural base.

Divergences in mandatory datapoints

The architectural consensus disappears when analysing the semantics of the data. The CIRPASS D2.1 document maps the existing legal requirements. Its Table 7 enumerates the requirements of Regulation 1007/2011 on textile labelling, requiring the fibre name and the percentage composition. The taxonomy accepted by CIRPASS inherits the classic denominations (acrylic, viscose, modal) and relies on voluntary initiatives to declare environmental impacts.

The JRC 145830 report disauthorises dependence on non-standardised schemes. In its section 2.2.2.2 on product-specific parameters, the JRC requires a direct link with Annex III of the ESPR. Steps C.1 and C.2 of its methodology impose the use of a "common cross-cutting semantic framework". Proprietary definitions or impact taxonomies inherited from isolated certifications lack validity if they do not map against recognised ontologies.

The requirement on the geographical origin of raw materials and the critical transformation processes exposes another fissure. CIRPASS D2.1 documents that the tracking of the deep supply chain (Tier 3 and Tier 4) is optional or fragmented in the current pilots. However, JRC 145830, through its data needs classification matrix (Table 15), evaluates the collection effort against the policy relevance. The JRC guidelines anticipate that the data on substances of concern (SoC) and circularity metrics will require rigorous controlled dictionaries. Textile brands face the challenge of migrating their ERP and PLM systems from free-text descriptors towards strictly validated JSON-LD properties.

Divergences in chain granularity

The granularity of the data determines the operational cost of the passport. Article 9(2)(d) of the ESPR Regulation allows the resolution to be set at model or batch level, and even to require tracking by individual item. CIRPASS D3.2 treats this decision as a configuration parameter of the resolver. The system architecture proposed by CIRPASS can route a batch identifier towards a DDR without prescribing business rules on when to force unit serialisation.

JRC 145830 closes this ambiguity. Its Annex 7, titled "Guidance on setting granularity levels", provides binding evaluation criteria for the legislator. The JRC establishes that if individual items accumulate disparate use or maintenance histories, "an item-level passport is justified" (JRC 145830, p. 104). The variability of characteristics and the accumulation of data in later stages determine the legal requirement.

For the textile sector, this methodological guideline is critical. A basic T-shirt without specific user data may justify a model or batch granularity. A high-value technical coat that includes a guaranteed repair service triggers the JRC criteria to require tracking at the unique-instance level. The divergence lies in the assignment methodology. CIRPASS prioritises current technical feasibility. The JRC designs the rules to force the shift towards unit traceability when the circularity metric requires it.

Risk of retroactive MVP invalidation

The divergence between the laxity of the early pilots and the rigidity of the JRC generates an imminent technical risk. Textile manufacturers that build a minimum viable model based exclusively on the CIRPASS D2.1 report may incur insurmountable technical debt. The main danger lies in the data governance structure during the product life cycle.

Annex 9 of JRC 145830 introduces the fundamental principle of the "Governed Life-cycle Data Framework". This model requires the technical separation between the "Core DPP" (static immutable data declared when placing the product on the market) and the "Life-cycle Log" (an append-only ledger to record subsequent events such as repairs and resales). The writes to this dynamic register require role-based permissions and cryptographic signatures linked to the actor's identity.

If a textile minimum viable model implements a monolithic JSON-LD document that mixes design attributes with repair updates through simple PUT calls to an API, the system will be retroactively invalidated. The 2027 textile delegated act, drafted under the JRC's rules, will require structured presentation credentials and logically separated life-cycle events. Furthermore, the recent Proposal COM(2025) 504 on digitalisation by default prohibits paper alternatives for compliance documentation. Brands will not have a grace period based on printed manuals if their digital infrastructure fails the semantic validation of the central European portal.

Decisional analytical reflection

The textile ecosystem is in a phase of regulatory transition. The network architecture is stabilised thanks to the work of the W3C and CEN-CENELEC committees. The variability lies in the ontological validation rules and in the granularity triggers. Operating under the ESPR umbrella requires anticipating the decisions of the delegated act by adopting the most restrictive methodology available.

The risk is not theoretical: a model built on the wrong document is invalidated with the 2027 delegated act and forces the documentary chain to be redone from the raw material supplier. The defensive decision is to start from the strictest framework available — the separation between an immutable core and dynamic circularity records of JRC 145830 — from day one, instead of discovering the technical debt with an entire catalogue already issued. It is imperative to decouple the internal attribute dictionaries from the public JSON-LD schemas. Consult the CIRPASS tutorial for building the first textile DPP-d21-schema-tutorial) to adjust the endpoints. Readjust your credential issuance flows by evaluating the analysis of Proposal COM(2025) 504 on digitalisation by default. Finally, check the technical analysis of the ESPR corrigenda R01 and R02 to align the operational compliance deadlines. The resilience of the digital passport depends on its capacity to absorb the imminent semantic precision of the legislator.