Understanding the ISO 15926 RDL

Information exchange with ISO 15926 uses a generic data model (which we call Part 2) that requires reference data (from Part 4) for it to make sense. For instance, one of the patterns in Part 2 is a generic pattern for a physical property. We use classes from Part 4 to make it a particular kind of physical property. So if we use a “temperature” class the pattern becomes a temperature property and if we use a “pressure” class it becomes a pressure property. But to make this work we need an accurate library of reference data that we can count on. We call this library the Reference Data Library, or just “RDL”.

The natural question is “What is this RDL?

The answer can be a bit confusing to new people because the term is used in two ways.

  • The first is an individual sense, as in “This endpoint exposes an RDL.” In this sense, an RDL is a single repository of information at a particular location, or endpoint.
  • The second is a collective sense. For instance, in discussions about ISO 15926 we talk about The RDL as if it were a single, specially consecrated repository managed by high priests somewhere in the mountains of Tibet. In reality, the RDL consists of many individual RDL’s distributed around the world. However, when we talk about the nature of the information in these individual RDLs and how it evolves, it is useful to talk about its properties as if it were a single entity, as we do below.

The RDL Contains:

  • The initial classes of Part 4
  • The initial templates of Part 7
  • Extensions to the initial classes and templates made by made by skilled modellers
  • Extensions to the initial classes and templates made by modellers of various skill levels

The following diagram has appeared, with slight variations, in many presentations, to show the federated structure of The RDL.

The yellow data base symbols represent individual RDLs in endpoints, individual locations for storing reference data. (For now, you can think of an endpoint as the electronic equivalent of a physical dictionary sitting on the shelf at a particular location.) They range from private sandboxes that are not open to the internet, all the way to the top level, which is published by the International Organization for Standardization (ISO).

Between the top and bottom levels is a continuum of endpoints serving different purposes, each with different levels of industry acceptance, or certification. Reference items, or terms, can move between these endpoints. Generally, the wider a term’s use, the higher it will reside on the continuum.

One issue is immutability. If you wish an information exchange to be reliable over time, one requirement is that each term used in the exchange must be immutable; that is, it must always exist and must always mean the same thing. Against the notion of “every term must be immutable” is the reality that some terminology genuinely has only a local meaning, or genuinely has one meaning over here but a different meaning over there. (And this is before we discuss human error.) The immutability of a term will emerge naturally due to the federated nature of the RDL. The lower level private sandboxes can each have a different meaning for a given term, and can change the definition of the term to suit the needs of their individual constituencies. As a consensus of the term’s meaning emerges, it can migrate up the certification scale.

Properties of the RDL

The RDL is worldwide

  • The RDL consists of all of the individual endpoints set up to serve ISO 15926 information exchanges.

No One Manages the RDL

  • Each endpoint is managed by its constituency. No one has overall responsibility for the combined content of all of the endpoints.
  • Endpoints are created for various purposes. They might be a private project-specific sandbox, or they might be created by an industry association and made available to its constituency or to the general public. When others find the content of a public endpoint to be reliable they will use it.
  • There is no requirement that any particular endpoint be used for any given information exchange. The partners of an information exchange will decide for themselves which to use. They may decide to use a particular set of endpoints for a set period, for a particular project, or forever. They can change their minds.
  • Everyone who looks at the RDL makes an individual decision on what they want to see. (But it is probably impossible to look at the entire RDL at once.)
  • If an endpoint contains information that, from past experience, is useful, use it. If an endpoint is published by an organization known for malicious or bad information, don’t.

Examples of Useful Endpoints

The RDL is Organic

  • The RDL is living and growing. As technology changes terminology emerges to describe it. From its inception, there was never the expectation that the RDL ever be complete.
  • The RDL moves. We have previously discussed the notion of varying levels of certification and that terminology can migrate between endpoints.
  • The RDL is imperfect. This is a good thing. It means that you don’t have to get a term analyzed and approved by anyone else, you can use it immediately.
  • The RDL can heal itself. No one forces anyone to use a term just because it is there. If a given term is wrong, use one from somewhere else.

An Endpoint Can be Made From Many Different Technologies

  • Microsoft Access
  • Oracle RDMS or Microsoft SQL
  • Spreadsheet

An endpoint can be on your own personal laptop (which likely means that it is not accessible to anyone else), or it might be hosted on a service that is open to the internet. Or anything in between.

A Given Information Exchange can Draw From Many Endpoints

The technology of the RDL is such that every term in even a very large information exchange can come from a different endpoint just as easily as if they all came from the same endpoint. (There might be a performance issue hitting thousands of endpoints instead of just a few, but the technology itself would work just fine.) Thus, the partners of an information exchange could decide to use top-level certified endpoints for 95% of the terminology, but use the endpoint of a niche industry for a few critical terms, and to use a project-specific endpoint for information that will be of no use once the project has been commissioned and turned over to the owner.

You Look at the RDL With Some Kind of Tool

  • Some tools, like the iRINGTools ReferenceData Editor, can see many endpoints at once.
  • Some can only see one endpoint at a time. Use whichever meets your needs.

The RDL was/is Created by People

Anyone can extend the RDL (even you); there are no privileges and no rights.

The people who have worked at the top level, ISO, have also worked on ad-hoc personal information. Each person who is now recognized as an expert went through a period where they were not an expert. The terminology they created during their learning curve sometimes reflects their level of expertise at that time.

There is no formal association of “official” RDL editors. It is a loose gathering of people who get to know and trust each other. If you want to join this group, just do it. Start contributing. As your expertise increases you will develop a sort of “street cred” and your work will be accepted.

Some of the RDL terms are ISO Certified. Generally, the ISO Certified information is of better quality, but being ISO Certified is not a guarantee.

Ideas on Using the RDL

Good Idea: Always use the highest level certification for each term in your information exchanges.

Bad Idea: Use a private endpoint for all the terms in your information exchanges.

Get out of the mindset of creating a private RDL endpoint for your organization’s information exchanges. Since most reference data is not unique to one organization, most of what you use should be from public, certified endpoints. Notice that we are not saying that the information exchange itself should be open to public scrutiny, only that the reference data be from public sources.

Think of the way we use commonly-accepted definitions of words to compose private documents. Even though a document may be highly confidential, it is made of words that come from the Oxford English Dictionary (or the equivalent for other languages). Only if the definition of a word needs to be clarified is it defined in the document. Likewise, highly confidential information exchanges can be made using reference material from public RDL endpoints. Terms where the definition itself is genuinely confidential or unique can come from a private RDL endpoint, but such terms make up a very small proportion of a typical document.

, ,

One Response to Understanding the ISO 15926 RDL

  1. Robin Benjamins July 16, 2012 at 6:49 pm #

    Gord,

    Excellent article on ISO 15926 reference data. This should provide non-technical understanding on perhaps the most important feature of ISO 15926.

    I have just some minor comments below:

    In the section “Examples of Useful Endpoints” I would modify the iRING endpoint to point to http://iringsandbox.org instead of http://iringug.org. This site has a long list of several RDL endpoints.

    In the section “An Endpoint Can be Made From Many Different Technologies” I would add to the list: RESTFul web services such as JORD, iRING etc.

    In the section “You Look at the RDL With Some Kind of Tool” I would add the commercial solutions: “XMpLant Technologies” from Noumenon and “Class Editor” from Bentley Systems, which also do federated reference data searches.

    Robin

Leave a Reply