Leeds BC Minutes

To enable inputs for interested parties in the US and Australia, the feasibility of Leeds-NIST and Leeds-Australia video conference links was considered. Regrettably, this could not be provided, as the existing LU video conferencing facility is too small (an upgrade is due soon). However, a number of email contributions have been received and these are reflected in this document.

Building Stuctural Frame: Steelwork (AP230) - Alastair Watson

Alastair reported that the CIMsteel Integration Standards (CIS) will be formally launched on 28th September which will give significant exposure to the CIMsteel work, and also to AP230. The first release of the CIS will included four Data Exchange Protocols sharing a common product model - they can be regarded as an ARM level AP with four conformance classes.

Apart from full project partners, CIMsteel has a significant number of associated collaborators. Recently signed associate collaboration agreements include Karlsruhe University and software house DSC from Germany. Alastair had also just visited Intergraph in Frankfurt and given a presentation to a Special Interest Group. Considerable interest in the CIS was expressed, particularly from another German software company. Wolfgang Haas has also been in contact with the German constructional steelwork association. At the time of the Washington meeting Alastair gave a presentation on the CIS and AP230 to representatives of the American Institute of Steel Construction (AISC) who have established a taskforce to participate in AP230. AISC will be hosting a second CIMsteel visit on 2/3 November to allow more details presentations on the CIS and AP230.

Leeds University has full time funding for around 2 years to work on AP230, mainly from the CIMsteel project in UK. Since Alan Boyle has now joined PDT Solutions, this post will be filled by Mike Ward (who attended the meeting and is due to start work on AP230 the 2nd October).

During the workshop, Alan led a walk through of the AP230 activity model originally introduced to STEP in Greenville. This was agreed subject to a number of minor corrections. Andy Crowley then led a walk through of the full CIMsteel data model which will provide the basis for the AP230 ARM. Paper copies of this model were given to those present. Electronic copies can now be distributed to others - anyone who wishes to contribute to AP230 please request a copy by email to Alastair or Mike (once he is in post).

Milestone dates for the work on AP230 are being set for Grenoble by which time it is hoped that a definitive version of the Activity model will be tabled. During the meeting copies of the following documents were tabled: AP230 Application Activity Model, A1 plot of CIS data model plus A4 CIS IDEF1X schema and powerpoint slides.

HVAC (AP228) - Jean-Luc Monceyron

Jean-Luc has now assumed responsibility for development of AP228. He is presently producing a detailed plan for the work which will be distributed by e-mail about 10 days before the Grenoble STEP meeting. HVAC should be a significant discussion at that meeting. Jean-Luc intends to refine the scope of the HVAC model at this workshop and to review the activity model initiated at the Nice workshop 12 months ago. Jean-Luc hopes to have a working draft for April 1996. The work is drawing on COMBINE, ATLAS and other existing partial models. Any other inputs or collaborations are welcomed. Felix Cheng advised of a UK project being conceived at present which would support the HVAC work. Wolfgang Haas has also advised the German association of building services engineers (VDI) of the AP activity.

INTRODUCTION

As from now, Jean-Luc will be the representative for CSTB (Sophia Antipolis). He will take charge of the AP on behalf of Patrice Poyet, current advocate. The HVAC AP is defined within the framework prepared under the auspices of the Building Construction Application Protocol Planning Project (BCAPPP) ; it will address the systems which provide the spaces of a building with heating, ventilating and cooling (for further details, see document ref. ISO TC184/SC4/WG3/N343 Proposal).

SCOPE REFINEMENT DISCUSSION

One of the main outcomes of the scope refinement discussion (to which Tony Baxter made a significant contribution), was that the AP should include piped-heating systems, and should meet the requirements of the majority of heating systems in Europe and elsewhere.
The scope remains as it was in the AP submission (not just ducted air systems).
Whereas maintenance and control were out of the scope of the proposal, there were requests to allow such information to be included or addressed in conjunction with process plant APs.
Concerning Product Data Typology, a more systematic approach will be further considered to order the expected data.

APPLICATION ACTIVITY MODEL (AAM)

Parts of the AAM from the previous draft were reviewed and discussed. There should be some outputs of the global activity at the top-level of the AAM these should be reflected down. Furthermore, there is a need of information in the HVAC AP considering the spaces within the spatial arrangement. A certain harmonisation could well be contemplated with BCCM afterwards.

PLANNING PROPOSAL

Ten days before the meeting in Grenoble, a draft including an activity model and a working plan should be released. It is noted that a working draft should be delivered as early as 31st March 1996.

Kent Reed reported (by email) intimate contact with ASHRAE. Members have seen the new work item but Kent is unable to motivate them to collaborate until there is a more formal workplan on the table. Kent also observed that making documents available only 10 days before an SC4 meeting means it is impossible for the US either to develop and forward a credible response or to arrange to send a representative to the meeting.

BC Core Model (Part 106) - Jeff Wix

Jeff reported that the BC Core Model was evolving satisfactorily. The objectives at this meeting are to get technical input and comment into some important aspects affecting the core model such as classification. Jeff has been invited to attend the ISO TC59 meeting in Stockholm next month, immediately before the STEP meeting. Anders indicated that there was some interest in the BCCM from TC59. Various models are being reviewed to find commonalties including AP225, AP217, AP221 etc. and those arising from projects COMBINE IDS, CIMsteel, ICON etc. Also models in EDI community (meetings with Mike Adcock). There is support from the UK DOE funded project for the work (managed by BRE) and the industry partners are being encouraged to say what was required to be incorporated. Jeff has already identified common concepts and been a 'thief in the night' by stealing them e.g. change requests from AP225 and some library concepts from Ship piping AP217. Jeff has had a meeting and telephone contact with Matthew West regarding the AEC Core Model concept. The AEC CM is likely to take the form of a higher order framework which positions and relates sector models. It could also provide an entity warehouse of common concepts to different sectors which they could adopt and specialise. At present no AEC CM development is taking place, effort and funding needs to be put in place.

Copies of extensive exploratory BCCM NIAM diagrams plus some supported text and definitions were tabled. During and following an extensive walk through the BCCM, the following points were raised in discussion with action required as indicated.

SCOPE AND PURPOSE

Development of the BCCM is progressing in three main areas.

(1) The development of a modelling framework which can be used by all model developments within the building construction sector.

(2) The inclusion of a number of models for areas of common interest which can then be used directly by other models and thus reduce modelling effort.

(3) The provision of an implementable aspect which can be used for the exchange of information between disciplines.

These areas of progress are intended to fulfil the requirements of Core Models as previously proposed by Matthew West (Shell) by (a) providing facilities which will reduce the time and effort required for AP development, and (b) providing a mechanism for the interpretation of APs within an industry sector in terms of its own terminology by allowing Application Reference Models (ARMs) to relate to the BCCM initially rather than having to undergo mapping directly to the Integrated Resources. Since the BCCM itself would be 'pre-integrated' with the Integrated Resources, interpretation of the AP to the BCCM at the ARM level would provide a direct interpretation for the AP.

These points were accepted and it was identified that the scope statement currently included within the BCCM should be expanded to cover these points. However, Alastair Watson suggested an action by Friday 29th September to produce a document which defined how each of the three points was being addressed (in order for email discussion to occur prior to Grenoble). This action was accepted.

DEFINITIONS

It was pointed out that definitions currently included in the BCCM Document have been completely reworked. Wherever possible, definitions are derived from the BS6100:1993 (Glossary of Building and Civil Engineering Terms) which is a superset of ISO 6707.

CLASSIFICATION

Anders Ekholm presented a series of slides on classification and indicated the interest of ISO TC59 in the development of product modelling. Classification is an unresolved area in the BCCM yet one which is considered to be important. To assist with liaison, Jeff has been invited to a meeting of TC59 in Stockholm in October. Prior to this, it is intended that an outline model for classification will be developed. Anders agreed to initiate and email conversation with Jeff to achieve this.

RESOURCE OBJECTS

Graham Cooper and Andy Crowley suggested that Resource examples included in BCCM at present should be removed. It was considered that the provision of examples in the definitions would be sufficient. This is to be considered as an issue.

USAGE SCENARIOS

Wolfgang Haas proposed that a number of usage scenarios should be developed in conjunction with industry. Jeff Wix indicated that this is already under way in conjunction with the UK CEIC project. Some of these will be ready for the Grenoble meeting

MULTI-FUNCTION SYSTEM OBJECTS

Graham Cooper suggested that the exclusion constraint be removed from the system objects so that an instance can be concurrently acting as more than one type of object. This is to be considered as an issue along with the 'acts-as' relation existing between various types of system objects.

SPATIAL ARRANGEMENT

A comparative review of the top level model in the proposed spatial arrangements AP was conducted. There was a close correlation between entity types. Jeff Wix to undertake a review and to recast this model using BCCM forms to demonstrate the correlation.

ENTITY COUNT

Robert Aish wondered if the BCCM might reduce its number of entities by splitting into several models using what he termed 'fault-lines'. Jeff replied that whilst some separation of the BCCM is not precluded in the future, there is no intention to do so at present since this would distort the BCCM from its present project orientation and run counter to the current understanding of what constitutes a core model.

LEAF MODELS

Jeff Wix explained that a leaf model is an outer part of the BCCM which is proposed for implementation as the interoperable area of data of common interest and at which APs would connect or map. A preliminary leaf model for spaces was suggested based on the COMBINE.IDM. This will be replaced by more specific reference to the proposed Spatial Arrangement AP. To further develop the leaf model concept, it was agreed that an exercise would be undertaken jointly with the CIMsteel model proposed as a base for AP230, and also with the HVAC AP. The results of this will be made available for the Grenoble meeting.

BUILDING AND SITE SYSTEMS

There was a general consensus that the provision of system objects for both building and site use was unnecessary and should be rationalised. This would assist in reducing the entity count of the BCCM.

LIBRARIES

The proposal library model was presented. This has been developed following there view of AP217. It was indicated that this is subject to further review by library specialists. No issues were raised against this model by the meeting.

Kent Reed asked (by email) when we will see a serious draft of the BCCM (on which the US can comment). Over 50 pages of BCCM material were distributed at the Leeds meeting, and I guess that Jeff would be willing to email this to interested parties. However, without a human guide, the practical value would be limited. Jeff was asked much the same question by those present - he plans to table a coherent version at Grenoble.

Building Elements by Explicit Shape Rep (AP225) - Wolfgang Haas

Wolfgang reported that after Washington meeting, Qualification had been carried out and the CD draft signed off on July 28. Balloting is now in progress, with closing date of 15th January. At end of July a first meeting of vendor test implementers had been held. Their first task was mapping tables between their products and AP225. These had been prepared and openly distributed to one another. There are now monthly meetings, and first test results are expected shortly. For the work, a toolbox had been developed for the vendors so that they effectively handled the problem at the ARM level, shielded from the AIM level by the toolbox. In response to a question from Alastair, Wolfgang felt that it would be possible for information on the toolbox to be made available, but not all problems were resolved and external people should hold back for the moment. With respect to the change request entity, none of the involved vendors had that capability in their software, though clients had expressed interest in having it. The Validation report (presently 100 pages) was well advanced (logging all meetings, issues, usage scenarios etc.). work on the Abstract Test Suite (ATS) had started, involving PDT Solutions, PDIT and Haas + P. Here the work was in new territory, since no ATSs had ever yet been qualified. A pre-print of an AP225 brochure was shown by Wolfgang (looked excellent in style) which should be available for Grenoble. Wolfgang had also made approaches to the German Concrete association and had made a presentation to them. They had agreed to pilot and implement AP225.

It is possible that the American Concrete Institute could be made interested in AP225 as well. We will know more as the AP225 ballot is circulated within the U.S.

Spatial Arrangements (planned AP) - Kjell Svensson

This planned activity will consider spaces and their spatial relationships with physical objects and support many activities from concept to operation of buildings. It was originally raised in Davos in 1994, and three different documents exist (4 page overview description of the proposed work item, an activity model in descriptive textual form, and a preliminary skeletal data model). There has been contact with Australia, and their ideas were being looked at. Anders Ekholm felt that ongoing research work at LTH could make a contribution to this work, as they are studying the early briefing stage of building projects. An e-mail conference was being proposed on this subject, and Anders would liaise with Kjell on this. The description of the new work item is expected be finalised for the Grenoble meeting.

The Spatial Arrangement AP was discussed mainly on Tuesday and Wednesday. Copies of the following documents were tabled:

A conventional description of the proposed new work item including a description of the scope and functional requirements.
An annex to the above document describing the needs of spatial information in different phases of the construction process.
A preliminary Spatial Arrangements AP data model (derived from the KBS model) and many other models within the same area.

The primary items discussed were:

SCOPE AND FUNCTIONAL REQUIREMENT

The functional scope of the AP is the description and representation of:

- the requirements on the internal and external spaces of the facility

- the spatial entities required to fulfil these requirements, and

- the physical entities that enclose and support the spatial entities.

The scope was discussed and mainly agreed, but it should be a bit more precisely described (for example what is not included in the scope). How will this AP interact with AP225, AP228, AP230, and Part 105 (with which it was seen as having a close and complementary role). It was noted that the AP would included both design (architectural) and end user (facilities management) viewpoints, their was some suggestion that these may need to be partitioned. The description of the proposed new work item should be finalised for the Grenoble meeting.

PRELIMINARY MODEL

The model was presented and discussed in some detail, particularly from the perspective of the BCCM. The linkage with the BCCM will be pursued in preparation for the Grenoble meeting. The discussions also highlighted the need for a complementary Building Fabric AP to carry detailed information about the building envelope. Members of the Australian AEC-STEP group sent an email describing their efforts and thoughts on the Spatial Arrangements AP. The material for Grenoble will be prepared by Sweden and Austria.

Roads (planned AP)

No report was available to the meeting, but planning is believed to be continuing.

Classification - Anders Ekholm

Lars Haggstrom is a member of ICIS and ISO TC59 SC13 and project leader for the BSAB96 system. Anders Ekholm, Lars and Kees Voesternek (NL) are working together on the information framework for classification within ISO TC59 SC13 WG2 and ICIS (International Construction Information Society). The ISO Technical Report "Classification of Information in the Construction Industry" was issued 1/7/94 as a type 2 draft and is now in review stage to be finalised in April 96. Representatives from BCCM and ISO TC59 are meeting in Stockholm in October to discuss common interests in establishing a construction information framework.

General Points

Graham Storer reported on interest shown by Mike Adcock in STEP modelling approaches. Mike is very involved in EDIFACT MD5 and had done some modelling of doorset specifications in ATLAS. The idea had been mooted of the product manufacturers (trade associations or interested individual companies) might develop data models of building product classes. These would be neutral, in the sense that they would be of use to EDI, Classification and STEP communities.
Graham also informed of a European project EU-BuildSTEP just submitted to EC to facilitate integration of the family of building APs, proving the core model concept, and involving vendors in prototype developments. Also, a UK project was to be submitted UK-BuildSTEP the role of which was to administer a network of UK researchers, industrial companies, IT vendors and research projects involved in product model developments (i.e. EPISTLE-like). BuildSTEP was prefixed EU- and UK- to enable other countries to use the name also if they so wished.
Following a presentation of the Steelwork Activity Model by Alan Boyle, it was agreed that there should be a 'generic' AM for the construction industry. Various models had been developed in projects, in STEP, in EDIFACT etc. but none had been pushed to acceptance and there had been no harmonisation between them. It was agreed that an approach should be made to ISO TC59 by Lars and Anders and to EDIFACT MD5 by Wolfgang to enquire whether there would be interest in a combined effort targeted at producing a proposed AM by June 1996. This then to go to the bodies for approval. Approach would be by at least two workshop sessions in the first half of next year involving say 8 industry people (2 structural engineers, 2 architects, 2 building services, 2 project managers from different countries) and perhaps 6 information modellers drawn from Classification, EDI and STEP teams. This might be an activity which could come under the umbrella of ELSEWISE, an already approved ESPRIT project beginning early next year and focused initially on business issues.
Concern was expressed about the new AP timetables, and exactly how they would be applied by STEP. Wolfgang reported that a First Draft was required within 6 months, and that a Committee Draft (including reference model, interpretation and qualification (now more difficult)) should be available after 2 years. The total process to final YES/NO DIS balloting had to be completed within 3 years. Wolfegang made the point that no current STEP AP had made this extremely tight schedule, which would need a committed team and the right people for each task. Alan Boyle tabled a page from a PDT Solutions proposal which gives there best understanding of the current situation (the key information is appended). It was agreed that this subject should be revisited in Grenoble, especially as BC work items such as AP230 would be badly effected by these new arrangements and will have the added complexity of the parallel development of the BCCM.
From the significant discussions about the planned Spatial Arrangements AP, two key points emerged. One was its close relationship with the BCCM and the HVAC AP. The other was the desirability of an early start on a Building Fabric AP (both to complement and to more sharply define the Spaces, Structural, and HVAC APs). It was agreed that the Grenoble agenda should include a revisitation of the APPP and the instigation of work on a Building Fabric AP.
There was some discussion of the Industry Foundation Classes (IFC). It was suggested that the phrase "conform to STEP" may mean that mappings from the IFC to the BC APs would be defined. It was noted that the IFC provides no control or resource objects and that objects such as walls, floors and windows are introduced at a very high level. Robert Aish was about to attend the first open Industry Association for Interoperability (IAI) meeting and he undertook provide a report to the Grenoble meeting. There was general agreement on the need for greater understanding of the workings of IAI, and of the need for technical co-operation.
Kent Reed made the following comment by email : "Most US AEC industry and academic people lack the support to attend standards meetings on a regular basis. As fallout from the political battle over the shape of the US federal budget, I expect a significant reduction even in my own ability either to go or to send others to foreign meetings. Workshops such as this one in Leeds will be even harder to get to than SC4 meetings. Furthermore, there is an ever increasing number of activities that need to intercommunicate. It is essential, therefore, that more BC materials be made available outside the meetings, preferably electronically, if the US is to make substantive contributions to the effort and to co-ordinate other activities. We can use either SOLIS or my own workstation facility to offer anonymous ftp and WWW access to STEP/AEC documents in general and BC documents in particular (I suggest my own facility only because I have direct control over the files and services, especially WWW) I could even offer a Lotus Notes server if there were enough demand for it. One way or another, we should be practising what we preach, namely collaborative engineering in a distributed environment using IT."

This was not discussed at the Leeds meeting, but should be raised in Grenoble. It is a problem that is not peculiar to the US (or to BC). We are now reaching the point were most of BC industry has email and WWW access, and we had hoped to have on-site email form the Leeds meeting. Unfortunately, the extension of the University network to the outlying "Devonshire Hall" was not completed in time. This document will be made available via SOLIS and WWW.

Toni Mury (tonim@tmabes.dialix.oz.au) provided the following email input (this was not read until after the meeting) : "I note that there still does not appear to be an AP group dealing with Electrical Services in buildings and in view of the development work we have carried out in this office in integrating the spatial and network models of electrical systems I would like express my interest in participating in any such group, should it eventuate. Our network model makes use of nodes, branches, routes, devices etc., together with their fundamental relationships, to represent any electrical power system. It is then used to carry out calculations, select components, and generate text and graphic reports of the system. We have been using InfoModeler, Informix and Access for the implementation and user interface, however John (Mitchell) is now encouraging us to move in the STEP direction. The spatial model locates electrical objects in real and classified spaces using conventional object based CAD systems (GDS, MicroGDS and AES). This is where our work interacts with that of both John Mitchell and Howard Leslie. I look forward to hopefully sharing some ideas with other like-minded parties."

Draft Grenoble Agenda

Sunday afternoon : BC Newcomers
Monday late morning : AEC Plenary
Monday early afternoon : BC Plenary
Monday mid afternoon : AP225
Monday late afternoon : AP230
Tuesday early morning : AP228
Tuesday late morning : Spaces AP
Tuesday early afternoon : IAI Report
Tuesday afternoon : Part 106
Wednesday morning : APPP Review and Fabric AP
Wednesday afternoon : Joint meeting (WG2)
Thursday morning : BC Plenary

Revised AP Project Timetables

Extract from PDT Solutions Document

Under the revised ISO/IEC Directives (Part 1 : Procedures, 1995), each project is required to meet the following target dates:

a working draft within six months of approval of the project;
a Committee Draft (CD) within two years of approval;
a Final Draft International Standard (FDIS) within three years of approval.

Further, the Migration Plan issued by the ISO Technical Management Board requires that:

all projects at the stage of working draft as of 1 July 1995 shall be required to advance to the stage of committee draft no later than 31 December 1996;
a first working draft for all work items approved prior to 30 September 1995 shall be required by no later than 31 March 1996.

Experience in the development and review of ISO 10303 Application Protocols has shown that a minimum period from the start of committee Draft ballot to the start of Final Draft International Standard ballot is at least 15 months. Using the terminology of the revised ISO/IEC procedures, this period is divided as follows:

Duration of Committee Draft ballot 4 months
resolution of CD ballot comments and preparation of DIS 3 months
registration and publication of DIS by ISO 1 month
duration of DIS ballot ("Enquiry stage") 5 months
resolution of DIS ballot comments and preparation of FDIS 2 months

September 1995, Leeds

Minutes Alastair Watson

This page is maintained by M A Ward

Last updated 23 April 1997

To the STEP Home Page at Leeds

To the AP230 Project Home Page

To the Civil Engineering Home Page

Minutes of the STEP WG3 / T12 / AEC/ Building Construction Sub-Group Workshop in Leeds 18-20 September 1995

Contents

Attendance

Apologies received included