One Assembly. Two Dataspaces. One Query.

How PMDCO (PMD Core Ontology) bridges incompatible industrial schemas across Catena-X and Manufacturing-X.

The integration problem

Catena-X and Manufacturing-X each expose material data through their own schemas — SAMM (Semantic Aspect Meta Model) aspect models for polymer properties, AAS (Asset Administration Shell) 3.0 inspection documents for steel certifications. A conventional integration requires a bespoke adapter for every pair of schemas.

The PMD approach is different: each dataspace's data is converted once into PMDCO (PMD Core Ontology)-conformant RDF, loaded into a shared triplestore, and from that point forward is queryable together using SPARQL — regardless of the original format.

The Tension Pulley Assembly demonstrates this: one physical part, two components sourced from incompatible dataspaces, unified by a single cross-dataspace SPARQL query.

CatX ──adapter──► MfgX
CatX ──adapter──► Lab
CatX ──adapter──► …
MfgX ──adapter──► ERP
…
N schemas × M targets
= N×M adapters

CatX ──► PMDCO ◄── MfgX
         ▲            ▲
        Lab           ERP
    one shared layer
    N queries possible
    0 additional adapters

Pipeline Simulator

Run each node individually — inspect the data it handles, then mark it complete. The knowledge graph inspector updates live.

Use the Run button in the Knowledge Graph State panel to execute the query. Load at least one source node first for results to appear.

SPARQL CONCAT aggregation reconstructs the original CatX JSON from the in-browser triplestore (urn:graph:catx-samm). The result is compared against the reference payload from the dataportal to verify the round-trip: CatX JSON → SAMM RDF → Oxigraph → SPARQL → JSON.

Full graph round-trip: store.dump({format:'application/ld+json', from_graph_name:…mfgx-aas}) serialises every triple from the named graph as expanded JSON-LD. The 12 expected values are located in the node array by AAS idShort and asserted against the original AAS inspection document (inspectiondocument_316_4401_alloy.json).