The OWL/RDF implementation of ISO 15926 consists of many interrelated and distributed modules, as is shown above. Each module has its own rate of change and related versioning. The data model, for example, has not changed since the year 2003, and project data at the other end of the spectrum change daily.
There are four main partitions:
and two auxilliary partitions:
The various modules are being discussed below. These are accessible by clicking the applicable symbol in the above diagram.
Please refer to the paper Redefinition of the Template Model for the underlying principles.
ISO 15926 OWL models
ISO 15926 starts with a data model, as defined in ISO 15926-2. In this ISO standard this data model is written in EXPRESS, the modelling language of the ISO 10303 ("STEP") suite of standards. See here.
For the purpose of ISO 15926 (Industrial automation systems and integration — Integration of lifecycle data for process plants including oil and gas production facilities) in Part 8 of that standard the choice has been made to use the technologies of the W3C Semantic Web.
The OWL version of the data model can be found at http://rds.posccaesar.org/2008/02/OWL/ISO-15926-2_2003 (Mozilla Firefox may give problems presenting that, IE does not).
A Turtle version can be found at http://15926.org/topics/data-model/owl-code.htm
In ISO 15926-3 a data model for geometry is defined. That model is an adapted OWL version of ISO 10303-42 that is written in EXPRESS, the modeling language of the ISO 10303 ("STEP") suite of standards.
The geo ontology imports the dm ontology.
The OWL version of the Part 3 geometry data model can be found here. (later)
ISO 15926 Parts 7 and 8 define the concept of Templates.
A ClassOfTemplate is a specialization of dm:ClassOfMultidimensionalObject and of dm:ClassOfClassOfInformationRepresentation.
A Template is a specialization of MultidimensionalObject and of ClassOfInformationRepresentation.
Both have a List structure.
A Template Model generically defines that List structure, with 2 - 11 predicates, and it defines the rdf:object of each predicate in terms of the attributes of dm:ClassOfMultidimensionalObject and MultidimensionalObject. This is fully described here.
Where the template model as such is actually used for reference only, this tm ontology is there for software developers to explicitly give role numbers to the roles of the template signature. The resulting code is made part of the OWL listing of the tpl ontology.
From the XML code of the Template Specifications the OWL listings are automatically generated and available at http://www.15926.org/templatespecs/__templates.owl.xml .
This tpl Template ontology also maps the template List structures from the tm ontology, and it defines the rdf:object of each predicate in terms of Part 2 and/or Part 3 and/or Part 4 specialized entity types, as defined in the Template Specifications.
This ontology contains TIPs, that are being discussed in the topic TIPs - Templates of Information Pattern .
Finally this ontology also contains some intersections like ClassOfActivity AND ClassOfInanimatePhysicalObject, or dm:ClassOfInformation Representation AND geo:Axis2Placement.
The tpl ontology is fully described here.
According ISO 15926-7, clause 5.1 and ISO 15926-8, clause 6.2 a specification for each template class and each specialization thereof shall be prepared, listing:
A set of approx. 180 template specifications for base templates has been posted here.
In the FOL listings, defined in the Template Specifications, use is made of Proto Templates as defined in Part 7 Annex C.
A typical example of a Proto Template is:
AssemblyOfIndividualTriple(x, y, z) ↔ AssemblyOfIndividual(x) ∧ ArrangementOfIndividualTriple(x, y, z)
AssemblyOfIndividualTemplate(y, z) ↔ ∃u(AssemblyOfIndividualTriple(u, y, z))
The full list is shown here.
In ISO 15926-6 an ontology for meta data has been introduced. In OWL terms these are instances of owl:AnnotationProperty. These meta data describe provenance data about objects and templates, not about their information content.
The Reference Data (in this definition) are, at present, mostly represented as a taxonomy. A definition for taxonomy is:
A taxonomy refers to either a hierarchical classification of things, or the principles underlying the classification. Almost anything, animate objects, inanimate objects, places, and events, may be classified according to some taxonomic scheme.
A typical example for a listing in the RDL (Reference Data Library) in RDF, according ISO 15926-8 Annex E, is (in Turtle):
Here specialized instances of the templates in the template ontology are stored. These templates refer to the reference concepts in the Part 4 Reference Taxonomy.
See here for an example.
Future Part 12
A group of OWL specialists in PCA (the POSC Caesar Association) is working on expressing ISO 15926 in native OWL, as was indicated in Part 8 Annex C. As it stands now ISO 15926 as described on this website is using OWL DL, be it for lowered templates. Part 12 will address the lifted templates.
Although this is too early to say, it would be appropriate in case the FOL would be mapped to that ISO15926-in-native-OWL-to-come. In that case it would be conceivable to use this ono ontology to define the rdl ontology, instead of the tpl ontology (that tpl ontology would be indirectly used anyway).
Here specialized instances of the templates in the template ontology in Part 12 format are stored. These templates refer to the reference concepts in the Part 4 Reference Taxonomy.
Extensions of the RDL defined by organizations
The Organization Reference Data Library is a combination of:
It is also possible to store translations of texts in a natural language other than English. These may stay in this Organization Ontology, or be offered to ISO for inclusion in the Reference Ontology.
Lifecycle Information and reference data
The Lifecycle Information store contains instances of entity types and templates in RDF format:
These instances are the result of mapping data residing in a COTS application, where only those data that originated there are being mapped.
Corporate Reference Ontology
The Organization Reference Data Library is mapped to OWL2 format to support reasoning. The degree of mapping may be determined by the objects that are in the discourse of reasoning.
Project data, mapped to OWL, for reasoning
Normally project data are being stored in RDF format, using SPARQL as query language. But in case OWL reasoning needs to be done, a "domain of discourse", a subset of the RDF data, is selected with SPARQL and mapped to OWL.