Re: Dataset Exchange Working Group UCR

Dear Jaroslav,

Now I have read the DCAT Recommendation I have come to the conclusion 
that it covers a world that is fairly remote from what we try to achieve 
with ISO 15926.

Having aid that, one day the two worlds will touch when you include:

  * our Reference Data Library <>, stored in a
    triple store
  * any life-cycle information about a facility of some kind

Then the DCAT rules apply.

Regards, Hans


On 16-3-2018 14:54, Jaroslav Pullmann wrote:
>   Dear Hans,
>     many thanks for drawing the ISO 15926 specification to our 
> attention and please excuse the belated response!
>    The standard seems to focus on a holistic modeling and exchange of 
> data in the process industries covering the entire
>    life-cycle of facilities. It would need a deeper understanding of 
> the standard to assess its implications on the more
>    general approach followed by the Data Catalog Vocabulary.
>    One of the shared concerns is apparently the evolution and 
> provenance of data that is currently being addressed by
>    the DCAT sub-group. Further insights into the requirements and 
> patterns underlying the ISO 15926 generic data model
>    would provide us with valuable hints to be considered in the 
> ongoing revision of the DCAT vocabulary. Please let us
>    know of your thoughts about the relation of both standards and the 
> potential ways how to utilize the experience gained
>    in designing ISO 15926.
>     Best regards,
>    Jaroslav (co-editor of the DCAT UCR document [1])
>  [1]
> On 25.02.2018 09:26, Hans Teijgeler wrote:
>> Hi,
>> Allow me to introduce ISO 15926, because I think there are some 
>> developments in it that may be of interest to you.
>> First I'll show you how the various parts of ISO 15926 fit together.
>> When, in 1991, we started with what is now ISO 15926, we did that 
>> because of the problem that everybody in the process industries has: 
>> how to exchange information without endless and error prone mapping.
>> In this industry most parties have rather "promiscuous" relations 
>> with each other. Fluor Corporation, in which I worked for 38 years, 
>> has some 2000 projects going at any point in time, ranging from very 
>> small (a one-man advice) to very large (billions of dollars). These 
>> are for many different owner/operators, in different countries, with, 
>> on very large projects, joint venture partners, and some 400 
>> suppliers and subcontractors.
>> All these parties have their own software with their own naming 
>> conventions, shortcuts, and logic. Even if they use the same software 
>> they have configured it differently to suit their work methods and 
>> procedures.
>> That is why we have designed a generic Upper Ontology (ISO 15926-2, 
>> see here <>), a Reference 
>> Data Library (see, for example, here 
>> <>) containing standardized 
>> instances of the ISO 15926-2 model, and in Part 7 templates 
>> <> that are constructs based 
>> on ISO 15926-2 and RDL classes. I guess that those templates are 
>> close to what you call 'datasets'.
>> To get a good feel of what Parts 7/8 entail please visit, as an 
>> example,
>> This means that we want to map and validate data that are produced by 
>> whatever application, used during the life of a facility, (and that 
>> means hundreds of applications!) _*at the source*_.
>> Those applications in most cases require information that was 
>> produced elsewhere (example: an app to size a pump by Mechanical 
>> Engineers requires process data produced by the Process Dept.).
>> Since ISO 15926 and its implementation methods guarantee integration 
>> of all life-cycle information, one can launch a SPARQL query in order 
>> to fetch the required up-to-date information.
>> Validation of the mappings is done following Part 10 (in 
>> development), using the new W3C SHACL <> 
>> Recommendation. A, yet inaccurate, description of our 
>> (work-in-progress) SHACL implementation can be found here 
>> <>.
>> Finally, the question is why anybody would like to store the entire 
>> history of his plant, from conceptual design through engineering, 
>> construction, testing, commissioning, operations (including thousands 
>> of 24/7 measurements over time frames of decades) and maintenance. 
>> Predictably it will become a treasure trove of knowledge about the 
>> plant, plant components, and processes.
>> In order to do such analysis the right hand top of above diagram 
>> shows that one can decide what the domain of discourse of analysis 
>> is, design a set of SPARQL queries to fetch all the necessary 
>> information, map this to OWL, and conduct reasoning sessions.
>> Needless to say that the above is a grand scheme. We spent a mere 
>> quarter of a century to get where we are, and are confident that in 
>> some years we will be able to roll out the first implementations.
>> Please note that ISO 15926 is NOT for application building, due to 
>> its generic (5NF) character such apps would run like a brick. ISO 
>> 15926 is for data management only:
>> and for an entire plant that can be pictured like this:
>> Finally this: in our industry the semantic preciseness of the 
>> exchanged information is of utmost importance due to the risks of 
>> lawsuits. That is why we had to approach the above with rigor.
>> I hope this synopsis was of some use to you.
>> Regards,
>> Hans
>> <>
>> ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ 
>> On 6-2-2018 18:05, Karen Coyle wrote:
>>> Note: This W3C working group will publish guidelines for application
>>> profiles. Use cases and requirements for this deliverable include
>>> interaction with, and possibly use of, validation languages.
>>> *** Please forward to potentially interested groups and individuals ***
>>> Hello,
>>> The Dataset Exchange Working Group (DXWG) [1] is pleased to announce 
>>> the
>>> publication of the First Public Working Draft of the Dataset Exchange
>>> Use Cases and Requirements.[2]
>>> The working group will produce a second version of the Data Catalog
>>> (DCAT) Vocabulary [3], guidance for the creation of application
>>> profiles, and content negotiation based on those profiles. The Use 
>>> Cases
>>> and Requirements cover all three deliverables.
>>> This document is the outcome of collaborative effort from the Working
>>> Group. We want to hear your comments on the document as it will guide
>>> the group in the three work areas. Please send any comments to the
>>> comments list [4].
>>> All feedback is welcome and will receive a response from the group. We
>>> look forward to hearing from you!
>>> The W3C Dataset Exchange Working Group
>>> --------
>>> [1]
>>> [2]
>>> [3]
>>> [4]
>> <> 
>> Virusvrij. 
>> <>
>> <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>

Deze e-mail is gecontroleerd op virussen door AVG.

Received on Friday, 16 March 2018 17:56:40 UTC