Changes between Version 1 and Version 2 of ISO15926Primer_HowItWorks_TechnicalDetails

11/20/11 03:38:42 (13 years ago)
gordonrachar (IP:



  • ISO15926Primer_HowItWorks_TechnicalDetails

    v1 v2  
    33= Technical Information About ISO 15926 = 
    5 ---- 
     5The '''ISO 15926 Primer''' has been replaced with '''An Introduction to ISO 15926''', a free download from Fiatech. 
    7 [[PageOutline(2-4,Contents,inline)]] 
     7This page is out of date and has been deprecated. 
    9 == Abstract == 
     9If you reached this page from a link in another web page please inform the webmaster. 
    11 ISO 15926 uses existing World Wide Web technology.  Tools developed for the Semantic Web have been leveraged to support the interoperability of plant information. 
     11For a peek at the new book and instructions on how to download a copy please follow this link. 
    13 ---- 
    15 == What ISO 15926 is For == 
    17 ISO 15926 can be applied to many different problems, but in practice today, it is a suite of standards that supports interoperability for data about the equipment and systems used in industrial processes, over the lifecycle of those objects. Some of typical scenarios for use of ISO 15926 are: sharing information between engineering contractors and their subcontractors, acquiring information from equipment vendors, handing over data to the operators of a new industrial plant and harmonizing the information across an enterprise that owns many plants. 
    19 == What ISO 15926 Standardizes == 
    21 ISO 15926 achieves its interoperability ends by standardizing all of the elements required for information transfer between systems: 
    23   * Part 2 - A temporal data model that defines and classifies information in space and time. 
    24   * Part 3 - Geometry and topology definitions. 
    25   * Part 4 - Reference data library for contributed classifications. 
    26   * Part 7 - A layer for concisely expressing relationships. 
    27   * Part 8 - Bindings to a serialization format for exchange. 
    28   * Part 9 - Bindings to protocols for query and transfer. 
    30 === ISO 15926 and The Semantic Web === 
    32 For the serialization format, part 8 defines bindings to OWL/RDF. The protocol for query defined by Part 9 is SPARQL - a language similar to SQL, but for ontologies in OWL/RDF rather than relational databases. SPARQL is built on WSDL, which in turn is built on SOAP using HTTP(S) as a transport. Similarly, the transfer and modification features are built on WSDL.  In short, as far as raw data interoperability is concerned, ISO 15926 relies wholly on the World Wide Web Consortium's Semantic Web standards (see, with of course the ubiquitous Internet Engineering Task Force standards HTTP and SSL underneath it all. These standards are well supported on every major enterprise platform: J2EE, .NET, PHP, and others in both commercial and open source implementations. The upshot is that in order to implement ISO 15926 all you need is one of these platforms and some semantic web toolsets. 
    34 == Retaining Interoperability with Customization == 
    36 New reference data agreed upon by communicating parties can be created and used as needed. These elements may be contributed back into the standard proper, simply made public (but not part of the standard), or formally marked as equivalent to other elements of the standard. This spectrum of approaches allows information management projects to gain effective communication in the short term and wider interoperability in the long term, without the delays attendant to forcing everything through the standards process from the start. What this means is that ISO 15926 can be rapidly adapted to the specifics of an integration project, without compromising the standard, or the portability of the data. 
    38 == Leveraging Protocols, Data Models == 
    40 The main benefits to an information technology project in using ISO 15926, rather developing from scratch, or developing new work over the top of OWL are that the hard work of defining the base temporal data model, serialization and protocol bindings has already been done. There are also consortia, such as FIATECH and POSC Caesar, which are dedicated to the public face and the public services of the standard - such as organizing for the central repository for reference data. In order to do what ISO 15926 does, all these parts would need to be developed. 
    42 == Distributing the Cost of Reference Data == 
    44 An object information model in ISO 15926 nomenclature loosely means the set of relations necessary to record information about an object of some particular classification. Formerly these were known as product models, but the name was changed since they apply to more than just products. Any system aiming to exchange information about industrial equipment and systems needs to have something equivalent to an object information model. 
    46 The work required to build an object information model for a typical piece of industrial plant equipment that is precise enough and flexible enough to form as a basis for interoperation is very high. The skills required to produce such models include some domain expertise in the relevant engineering fields along with data modeling expertise. Given that there are at the very least thousands of these classifications in a typical process plant such as a refinery, the cost for a software vendor to "go it alone" on this front is prohibitive. For a software services company, the risk of not acquiring sufficient people with the right skills is high, and even if it were to be done successfully, the cost passed on to the customer would be extremely high.  So the collaborative, standards-based approach to reference data allows all organizations using ISO 15926 to contribute small pieces, and benefit from the work of all the other organizations - distributing and reducing the total cost, and retaining domain knowledge across generations. The way that ISO 15926 works allows the object information modeling work to be pushed out to the software user, where there is already engineering domain expertise - engineering companies, plant owners, equipment vendors. The benefit of this approach to software vendors and services companies is near incalculable - it makes the difference between a feasible project and an infeasible project: between a favorable cost/benefit ratio and an unfavorable one. 
    48 == Service Oriented Architecture (SOA) == 
    50 One of the primary uses of ISO 15926 is to present an enterprise software product as a Service Oriented Architecture. The first and most obvious advantage to ISO 15926 as an SOA implementation is that it is standard: any software vendor can compete, and products can interoperate without infringing on legal rights or dealing with uncertainties of proprietary definitions. Also, since the standard provides the query and transfer layers, there is little required in terms of service definition. This is particularly attractive from the product planning point of view - it is much easier to encourage other vendors to operate against a product's SOA definition if that definition is also used in yet other vendors' products. 
    52 == Representational State Transfer (REST) == 
    54 The second, and possibly more important advantage, is that ISO 15926 operates on principles that are similar to Representational State Transfer (REST). Without going into great detail, the net gain of this approach is that rather than having a large number of methods or functions at the SOA API level, you have relatively few. The few available functions are designed for either querying data, or just for moving data from one system to another. They do not, (for the most part anyway), have functional side-effects. The great benefit this brings to SOA is that implementation of an endpoint is relatively easy - it does not require implementing a lot of discrete, algorithmic behavior behind a large number of methods. It is a substantial reduction in implementation risk. 
    56 == Staging ISO 15926 Implementation == 
    58 Many of the concrete gains of ISO 15926 can be realized without even necessarily engaging in part 7. For example, uniform classification enabled by ISO 15926 part 4 reference data can be an enormous gain to enterprise-wide data harmonization.  When moving into parts 7, 8, and 9 territory, OWL/RDF can be used for data import or export of ISO 15926 data (Part 8) without going as far as implementing façades (the SOA portion - Part 9) - this sort of solution could be used for generating or consuming handover data. Finally, once both import and export are implemented, the vendor could implement full SOA to expose real-time data held within their systems, automatically enabling ad-hoc queries and taut integration with other applications and services. 
    60 == Conclusion == 
    62 ISO 15926 brings tangible benefits to enterprise software in the process industry realm. As a suite of layered standards, it allows implementation to be staged into individually useful pieces, and therefore, allows risks to be managed. The standards' utilization of the W3C's Semantic Web as a foundation for interoperation ensures toolset availability and relevance now and into the future. Prominent software vendors with traditional markets on both the engineering and plant operator sides are investing substantially in this standard today. Engineering, Procurement and Construction companies have long supported the initiative with the expensive secondment of staff and resources. Plant Owner Operators are writing ISO 15926 into contract terms. ISO 15926 is a market opportunity. 
    64 == Next == 
    66   * [wiki:ISO15926Primer_Implementations_WhatBeenDone Primer: What Has Been Done to Develop ISO 15926?] 
    68 === Acknowledgements === 
    70 Thanks to NRX Global for the information this page is based on. 
    72 ---- 
     13  * [wiki:ISO15926Primer An Introduction to ISO 15926] 
About PCA
Reference Data Services