ISO TC184/SC4/WG10 N316
2000-09-26
Source: editor
Julian Fowler, PDT Solutions (editor)
Note - these minutes including all attachments are also available as a single ZIP archive.
The agenda for the meeting was confirmed as follows:
Participants in the meeting were:
| Matthew West | Shell | UK | |
| Hiroyuki Hiraoka | Chuo University | Japan | |
| John Nazemetz | Oklahoma State University | USA | |
| David Price | IBM | USA | |
| Julian Fowler | PDT Solutions | UK | |
| Rogerio Barra | ATI/PDES, Inc. | USA | (item 4.2 only) |
| Steve Chilcott | Rolls-Royce | UK | (item 4.5 only) |
A number of PDES, Inc. member company representatives also took part in the "STEP On The Web" discussion.
Issues raised prior to the meeting by Julian Fowler (see Attachment A) were reviewed and discussed.
Issue 4 - need to reflect that being an integration model is a role.
Issue 6 - leave this issue open - are we using "map" and "mapping" with single consistent meanings throughout the ISO 18876 documents?
Issue 15 - need to separate out the principles that:
Issues raised by Chris Angus prior to the meeting were also discussed - these were included in an annotated MS Word document provided to Matthew West. Most of the changes were editorial in nature. It was agreed to accept Chris's proposed change to the use of the words "map" and mapping.
The overlap between Part 1 and Part 2 was discussed, noting that the level of detail of the overview of mapping and integration in Part 1 is not dissimilar from that of Part 2. Julian Fowler noted that this may not be so much of a problem, since development of Part 2 has demonstrated that a description of the methodology that is not specific to a modelling language or a given integration data model. Matthew West suggested that Part 2 should provide this additional detail for models that are defined in EXPRESS. This was agreed as a refinement of the scope of Part 2. Matthew also suggested that the material in clause 5 of Part 1 could be repeated in the Introduction to Part 2.
Julian Fowler suggested that both Part 1 and Part 2 should include statements (in their Introductions) of the purpose and the target audience for each Part.
The text of Part 1 was further reviewed in detail and a number of changes made to the document. It was agreed to replace the terms "integration data model" and "application data model" by "integration model" and "application model" throughout.
Matthew West stated that he intends to produce a second working draft of Part 1 for distribution approximately one week prior to the Charleston meeting.
The first working draft of ISO 18876-2 (WG10 N315) was reviewed in detail. It was agreed that the structure of the document should be updated to reflect two major subdivisions:
It was also agreed that within the subdivisions, the document should cover:
The nature of an integration model was discussed, with the following conclusions:
It was agreed to delete references to data exchange, sharing, etc. from Part 2 - these can be discussed in Part 1 but are not issues for integration models.
The following additional issues were noted and discussed:
Matthew West suggested that Jim Fulton's paper "Semantic Plug & Play" should be circulated as a WG10 N-numbered document, and should be cited in the Bibliography of Part 2 as useful background reading.
Julian Fowler stated that a revised and updated draft of Part 2 will be available for review shortly before the Charleston meeting.
Matthew West presented slides prepared by Chris Angus on the relationship between EXIST and KIF (see Attachment B). Julian Fowler noted that potential use of KIF has been discussed on the IEEE's discussion group on Standard Upper Ontologies (SUO). Jim Fulton (Boeing, and a former member of WG10) has been working on an XML encoding of KIF. Chris Angus's recommendation that KIF should be regarded as a possible interchange format for EXIST models was accepted.
Matthew West led a discussion on refinements to an EXPRESS integration model, using WG10 N220 and ISO 15926-2 as starting points.
A number of subject areas were identified as being required for a general-purpose integration model:
A number of refinements and improvements were made to the integration model and were documented in a series of EXPRESS-G diagrams - see Attachment C.
Using the integration model components developed in the previous part of the meeting, a start was made on mapping the "product identification" module developed by PDES, Inc. to the integration model.
The precise nature of a mapping was discussed, leading to a proposed use of formal English language statements to characterize a mapping, as shown below:
For each x
(where x is a member of Constrained_parts_list.Product
Then x is a member of class
...
For each z (where
z is a member of Constrained_parts_list.Product.ID,
There exists some y,
(y is a member of Constrained_parts_list.STRING,
z=<x,y>
Then:
There exists some a, b, c, d, e
(Where:
a is a member of IM.EXPRESS_string,
b is a member of IM.EXPRESS_string.content,
c is a member of IM.class_of_reference_by_encoded_information,
d is a member of IM.class_of_reference_by_encoded_information.referenced,
e is a member of IM.class_of_reference_by_encoded_information.referencing,
and
b=<a,y> ,
d=<c,x>,
e=<c,a>.
)))
)
During the development of this mapping the importance was recognized of:
A sample instance mapping diagram is shown below.
It was agreed that this needs to be documented as "good practice" in ISO 18876-2.
WG10 members met together with PDES, Inc. participants to discuss the topic of "STEP On The Web". Julian Fowler noted the background to WG10's interest in this area, explaining that the working group has an outstanding action to produce an SC4 white paper on this topic. The nominated editor of this paper is Bill Burkett (P.D.I.T., Inc., USA).
David Price presented an overview of PDES, Inc.'s interest in "STEP For The Web" He then presented the current status of ISO 10303-28 "XML representation of EXPRESS schemas and data", focusing on:
Joe Trausch (Electric Boat) summarized the ISE (Integrated Shipbuilding Environment) project, and stated that EB is interested in both the object serialization early binding (OSEB) and containment early binding (CEB) in Part 28, and mappings between them. EB's goal is to move towards the use of XML Schema rather than DTDs for validation of XML data. The ISE project also includes parts catalogue work related to ISO 13584 PLIB.
David Price presented some additional information on the OSEB (see Attachment D), focusing on the comparison and overlaps between:
A short discussion followed on the definition and use of Enterprise Data Objects (edo) within the OSEB, and the use of simple and extended XLinks as part of the Part 28 specification.
David Price used slides provided by Martin Harwick (STEP Tools, Inc.) to summarize the containment early binding (CEB), which is targetted for inclusion in edition 2 of Part 28 (see Attachment E).
Dave Campbell (Northrop Grumman Logicon) reported on a trial implementation of the Part 28 late binding, intended to test the applicability of Part 28 for PDM to a supplier product structure delivery system).
David Briggs (Boeing) reported on an AP 224 feature based machining project that has investigated XML as an alternative to Part 21. The XML representation is then used as the basis for identifing part features for use in a cost estimating system.
David Price presented IBM input (see Attachment F) that provides rationale for the development of Part 25 (XMI binding). The goal of this work is to make EXPRESS schemas available - by mapping EXRESS to UML and the XMI DTD - to users of UML-based modelling and implementation tools. Further work will investigate possible convergence between the OSEB and UML.
David Price then closed the session by presenting a short paper prepared by the PDES, Inc. STEP For The Web team (see Attachment G). This is available to WG10 as an input to the working group's task in this area. PDES, Inc. has established a website (http://www.stepml.org) to publish information about the use of STEP in conjunction with XML and other web technologies, including usage guides for XML DTDs derived from AP schemas.
The following actions were noted from this session:
Possible Action Items:David Price's slides presenting the agenda for this part of the meeting and the current status of the STEP Modularization PWI's deliverables is available as Attachment H.
The resolution passed by SC4 in Bordeaux about the creation of a Modules Harmonization Team was reviewed. The SC4 requirement for published guidelines is interpreted as covering the following documents:
It was agreed that these three documents will be submitted for standing document ballot. The procedure for doing this (specified in the SC4 Quality Manual) was reviewed:
David Price stated that he would aim to submit new versions of all three guidelines documents for WG10 review by 2000-09-30 in order to meet SC4's deadline of publication by 2000-10-31.
Rogerio Barra presented PDES, Inc.'s proposals for standardization of further modules in the PDM and engineering analysis domains. Some 40 modules are currently being prepared. Rather than submitting these for 40 separate NWIP/PDTS ballots, it is proposed to group these into a smaller number of "packages" so that P-members have a smaller number of votes to consider.
Julian Fowler asked whether grouping of several modules into a single part document had been considered? Rogerio replied that such a change would cause a great deal of redocumentation work. Julian Fowler noted that although there is no precedence in SC4 for considering multiple parts in one stage 30 (CD/PDTS) ballot, this has been done in the past for NWIPs (e.g., ISO 15926 which was approved as a single NWIP vote on three parts).
It was agreed that the proposed approach is a useful one, and that David Price should prepare a last-day resolution for consideration by SC4 in Charleston. This resolution will enable balloting of "packages" of modules on a trial basis - if the resolution is approved and the "package balloting" proves successful, the SC4 Handbook can be modified later to enable this as a general practice.
The previous version of the AM Content Guidelines (v0.6, WG10 N220) was reviewed. A number of modifications to the document were discussed and agreed. Issues noted included the following:
David Price will update the document for review by WG10 - the revised document will be published as WG10 N317.
The previous version of the Modular AP Content Guidelines (v0.6, WG10 N222) was reviewed. A number of minor modifications to the document were discussed and agreed. David Price will update the document for review by WG10 - the revised document will be published as WG10 N318.
The previous version of the AP/AM Development Process Guidelines (v0.2, WG10 N223) was reviewed. A number of modifications to the document were discussed and agreed. Issues noted included the following:
David Price will update the document for review by WG10 - the revised document will be published as WG10 N319.
The following proposed sessions in Charleston were noted:
The full schedule and agenda for WG10's meetings in Charleston are available on the USPro and PDT Solutions websites.
It was agreed that there will be a WG10 workshop between the main ISO meetings in Charleston and Portugal. The meeting will be held at Shell Centre in London, provisionally from 2001-01-15/19 (dates to be confirmed in Charleston). If needed, an ISO 18876 editing meeting will be arranged for the start of the following week (2001-01-22/23).
All participants thanked PDES, Inc. for hosting the meeting.
Except where noted, all attachments are Adobe Acrobat (PDF) files.
Respectfully submitted
Julian Fowler
2000-09-26
(with thanks to David Price, John Nazemetz, and Matthew West for the use of their notes as the basis for parts of these minutes)