News

About the project

Our PASSTEX project, numbered 3260667 and submitted to the first 2026 call of the 1501 Industrial R&D Support Programme run by the TÜBİTAK Technology and Innovation Funding Programmes Directorate (TEYDEB), was found worthy of support following evaluation. The funding period runs from 1 July 2026 to 31 December 2027.

Full project title: PASSTEX – An AI-Powered Digital Product Passport Platform for a Sustainable and Circular Textile Ecosystem.

Why a Digital Product Passport?

With Regulation (EU) 2024/1781 on Ecodesign for Sustainable Products (ESPR), the European Union placed the Digital Product Passport within a binding legal framework and positioned textile and apparel products among the priority product groups of its 2025–2030 working plan. The regulation obliges manufacturers shipping products to the EU market to present verifiable data covering product identity, fibre composition, chemical compliance, origin, carbon and water footprint, recycled content share, and repairability and recovery potential.

For the textile sector, one of Türkiye's most important export categories, this obligation is not merely a technical compliance question; it bears directly on market access. Sector data, however, is mostly held in fragments across enterprise resource planning systems, manufacturing execution systems, laboratory results, supplier declarations and document-based manual processes. That fragmentation raises both compliance risk and operational cost.

The PASSTEX approach

PASSTEX treats the digital product passport not as a single-stage document solution but as a phased, modular product governance infrastructure that adapts as the regulation matures. The platform's original R&D contribution lies in developing three components together.

Semantic data mapping

Heterogeneous data arriving from different points in the supply chain is mapped onto a common semantic layer through a textile-specific ontology. The extensible JSON-LD based data model integrates with ERP and PLM systems through standard API connectors, and can communicate with the EU Data Spaces ecosystem through protocols such as NGSI-LD.

Explainable compliance engine

Large language models turn international regulatory texts into machine-readable rules; those rules are modelled in a semantically queryable knowledge graph; and machine learning models produce anomaly detection and dynamic ESG scoring. The three components combine in a hybrid compliance engine that performs both pre-production scenario analysis and post-production verification. Through an explainable AI layer, the system shows auditors transparently which data fields and which chain of rules each decision rests on.

Verifiable trust layer

Building on the W3C Decentralised Identifiers (DID) and Verifiable Credentials (VC) standards, the project is developing an infrastructure in which every stakeholder in the supply chain manages its own cryptographic identity. Industry certifications such as GOTS and Oeko-Tex, together with sustainability declarations, are turned into verifiable digital credentials. Assurance levels that differ by data class deliver a low-cost architecture matched to risk and audit requirements, rather than a uniform and energy-intensive one.

Who is it for?

The platform targets users at two levels of maturity: large-scale manufacturers running their own enterprise IT infrastructure, and small and medium-sized enterprises with limited digital maturity and no IT infrastructure of their own. The architecture, deployment model, integration strategy and user experience are differentiated for these two segments.

Web-based management dashboards will be developed for manufacturers and brands, and a mobile application giving access to a product's sustainability story through a QR code will be developed for consumers and for verifiers in the second-hand market.

Project timeline

The 18-month funding period is divided into three monitoring periods. The first covers requirements analysis, data architecture and core platform infrastructure; the second, the interoperability layer, the decentralised trust infrastructure and the AI components; and the third, user interfaces, the integrated AI engine, system integration and pilot validation. The project will conclude with end-to-end scenario tests run on real production data at pilot textile manufacturers.

Acknowledgements

We thank TÜBİTAK for supporting our project, Biruni Technopark for its hosting and ecosystem support, the Bahçeşehir University Technology Transfer Office for its academic guidance, and our pilot business partners for their collaborative approach.