GS1 in Europe experiments with ADMS

Douglas Hill has spent over thirty years in supply chain management but in the last sixteen he has specialised in Electronic Data Interchange standards and standardisation in the supply chain. In his current position as COO (eBusiness) for GS1 Denmark he advises major Danish FMCG, pharmaceuticals and other sectors on E-Business solutions. GS1 in Europe eDox platform is expected to join the ADMS-enabled federation of semantic asset repositories in June 2012.

Joinup team: Which benefits do you expect from the federation?

Douglas Hill: The eDox platform would likely be designed differently, if the ADMS specification were available at the time of its design and development. GS1 in Europe wants to federate some of the content in its eDox platform to give more visibility to its work and that of its Member Organisations. In time, GS1 in Europe considers better aligning its internal data model with ADMS, for example at the level of the taxonomies.

Joinup team: What are your experiences with mapping to ADMS?

Douglas Hill: The mappings can be done at two levels: At concept level and at the level of the controlled vocabularies.

At concept level, we really needed to align with what the ADMS specification considers to be semantic interoperability assets. Some documents in the ADMS specification are documents about semantic assets, rather than interoperability assets in their own right. With regard to granularity, we would like to have an asset release with a ZIP file format. GS1 in Europe has many document types related to a standard, for instance the Business Messaging Standard for e-Invoicing, is packaged as a syntax guideline with accompanying documentation such as data models and examples. However, we cannot easily and automatically separate these documents out, and actually, neither do we want to. The idea at this messaging guideline level is to give the user an implementation package.

At the level of the controlled vocabularies, asset type is a foundational vocabulary, which currently is still under debate in the ADMS Working Group. Other controlled vocabularies are easier to map to. For example, I was able to provide a draft mapping for the sector taxonomy in the eDox system to the top-level EuroVoc domains. When exporting to ADMS, we could easily cater for these mappings.

Joinup team: How will you export the GS1 metadata?

Douglas Hill: At this point in time, it is not yet clear how many assets from the eDox platform will be exposed to the federation but we expect about 80% to be made available. At this time there are 386 registered standards and documents in the eDox repository, where most are publicly accessible without the need for any special access rights. In the longer term, the GS1 Global Office might consider sharing the GS1 Global Data Dictionary with the federation, where each contained core component can be viewed as a separate asset. Currently GS1 has about 3,500 context free core components on which it has built its business messaging. These assets are fully compliant to ISO/TS 15000-5:2005 Core Components Technical Specification (CCTS) and form the foundation of the GS1 Global Data Dictionary.

About the export method, my opinion is that the eDox database will first be exported in a proprietary XML schema. The developers and system administrators are based in Hungary.

Joinup team: How will the description metadata be exchanged with Joinup?

Douglas Hill: I expect that the first exchange of metadata would probably take place via e-mail (solution 2) or through manual file upload (solution 4). After the pilot, metadata exchanges could be automated using a web service client.

Any of the global GS1 organisations can upload its specifications, so changes in the eDox repository can happen any time. However for the first phase, a weekly update of the metadata seems reasonable.