Master's Thesis — SE4GD Erasmus Mundus

Interoperability Platform for Digital Product Passports

A Design Science Research study using Linked Data technologies to bridge heterogeneous DPP initiatives and enable cross-initiative semantic queries.

Jorge Renato Leon Chumpitaz — February 2026

Click chapters to expand
01

Introduction

The European Green Deal and the Circular Economy Action Plan mandate machine-readable Digital Product Passports (DPPs) by February 2027. Multiple initiatives — Catena-X / Tractus-X, the Global Battery Alliance Battery Pass, and CIRPASS — are independently creating overlapping but incompatible schemas. This thesis addresses the interoperability gap with a platform that resolves heterogeneous identifiers and normalizes payloads into a shared semantic model.

EU Battery RegulationDigital Product PassportInteroperabilityResearch Questions
02

Background

This chapter surveys the regulatory landscape (ESPR, EU Battery Regulation), the major DPP initiatives, and the Semantic Web technologies underpinning linked data — RDF, OWL, JSON-LD, SPARQL — and how they enable cross-initiative queries. Design Science Research (DSR) is introduced as the methodological framework.

ESPRSemantic WebJSON-LDRDFDesign Science Research
03

Related Work

A systematic review of existing interoperability approaches: schema mapping studies, identifier resolution services, DPP prototypes from industry and academia, and Linked Data Platform architectures. We identify five specific gaps that motivate the design of our artifacts.

Schema MappingIdentifier ResolutionLinked Data PlatformLiterature Review
04

System Overview

The DPP Interoperability Platform comprises two microservices — an Identifier Resolution Service and a DPP Normalizer — linked to a semantic triple store. The architecture follows a product-agnostic core + modular extension design (v0.2) with JSON-LD as the canonical serialization format.

MicroservicesArchitectureProduct-Agnostic CoreExtensions
05

Design

The Shared Core Schema defines 17 product-agnostic fields plus three extension schemas (battery, state-of-charge, LCA/environmental footprint). Identifier resolution uses a five-type taxonomy with multi-registry lookup and confidence scoring. Every normalized field carries PROV-O provenance metadata.

Shared Core SchemaPROV-OProvenanceExtension Schemas
06

Implementation

The complete TypeScript/Node.js implementation: 4,626 lines of code, 331 passing tests, four adapters (Tractus-X JSON/JSON-LD, Battery Pass JSON/JSON-LD), OpenAPI 3.1.0 specs, and the Next.js interactive UI. All services produce dual-format outputs (REST JSON + JSON-LD/RDF).

TypeScriptVitestOpenAPIAdaptersNext.js
07

Evaluation

Functional verification through 331 automated tests, cross-initiative SPARQL queries demonstrating 5.4x code reduction vs. REST, TriplyDB dataset validation with approximately 900 RDF triples across 7 passports, and practitioner interview preparation for Phase 1 verification.

TestingSPARQLTriplyDBCode ReductionValidation
08

Discussion

Analysis of how the platform addresses each research question RQ2-RQ5, the trade-offs in schema design, JSON-LD vs. Turtle serialization rationale, lessons learned from adapter development, and implications for the broader DPP ecosystem.

Research QuestionsTrade-offsJSON-LDLessons Learned
09

Threats to Validity

Internal threats (schema completeness, mock data reliance), external threats (limited to two initiatives), construct threats (coverage metric validity), and reliability considerations (reproducibility via open-source artifacts).

Internal ValidityExternal ValidityConstruct ValidityReliability
10

Conclusion

The DPP Interoperability Platform demonstrates that Linked Data technologies — JSON-LD, SPARQL, and a modular shared schema — can bridge the gap between independently developed DPP initiatives. Future work includes CIRPASS onboarding, practitioner evaluation, and production hardening.

ContributionsFuture WorkCircular EconomyOpen Source

Explore the Platform

See the research artifacts in action — resolve identifiers, normalize passports, and query the semantic graph.