BIModellers can share little or much information available across desperate industry domains. The optimal BIModeller would have the ability to display, calculate and share all data necessary between disciplines without loss or workflow conflicts This ability, or lack of, is a function of the technology used, the process deployed and the parties (knowledge workers) involved.
This episode is available in other languages. For a list of all translated episodes, pleaser refer to http://www.bimthinkspace.com/translations.html. The original English version continues below:
Assuming each domain (industry sector: Architect, Engineer or Constructor) is using a different BIModeller, data sharing methodologies amongst these modeller can take many forms:
- Data Exchange: Each BIModeller keeps its integrity but exports out some of its ‘sharable’ data in a format that other BIModellers can import and calculate (think XML, CSV or DGN for example). This method is arguably the primordial data sharing method and suffers from the highest unintentional data loss rates. Data loss here signifies the amount of data that can’t be shared as compared to overall data available in BIModels. However, not all data must or need to be shared between BIModellers all the time. Partial Data Exchange (as compared to unintentional data loss) can be an intentional and efficient method of data sharing [1].
- Data Interoperability: Interoperability can be in many forms; the one discussed here is merely an example. Assuming file-based data interoperability (not server based interoperability) one of the demonstrated scenarios for this data sharing methodology is as follows: BIModeller 1 produces an IModel (Interoperable Model) that gets imported into BIModeller 2 where its worked-on then exported into IModel v2 (version 2) that gets imported into BIModeller 3 where its worked on then exported into IModel v3 that gets imported into…[2] The amount of data lost/gained between modellers, models and model versions depend on modellers’ import/export abilities and the interoperability schema itself (think IFC or CIS/2 for example). One major shortcoming of this ‘file-based’ interoperability is workflow linearity; the inability to allow simultaneous interdisciplinary changes to the shared.
- Data Federation: File linking is a good example of data federation: data in one BIModel is linked to data in another BIModel. The files are neither imported nor exported but BIModellers (software applications) can read and calculate the data embedded within the linked files. The amount of data loss depends on the amount of data readable or calculable. Referential Models (RModels) are another example of BIM Data Federation. RModels are single or federated models that host links to outside data repositories; much like a hyperlinks on a webpage. An example of this would be a virtual building with a referential window object: detailed information (values) beyond the basic parametres are not saved within the BIModel but are accessed from an external repository whenever the need arises [3] (ex: realtime window cost, availability, installation manual, maintenance schedule).
- Data Integration: The term integration may be understood in many ways including the lower-grade ability to exchange data between software solutions. In a BIM context, an integrated database signifies the ability to share information between different industry sectors using a common model [4]. The sharable data within the BIModel may be architectural, analytical (engineering) or managerial as well as design, cost or code information (refer to Episode 4). What is important about an Integrated BIModel is that it co-locates interdisciplinary information allowing them to interact with each other within a single computational framework. At this moment in time, few if any available BIModellers can integrate the necessary data and processes needed to achieve an integrated model (A model of a set of data is an interpretation under which they are all true [5].
- Data Sharing Hybrid: A combination of any of the data sharing forms discussed above. Most BIModellers, whether proprietary or not, coordinate the multi-disciplinary information generated by AEC sectors through hybrid of information sharing methodologies.
Figure 5.1: Data Sharing Forms
References:
[1] Fischer, M. and Kam, C. (2002) PM4D Final Report, CIFE, Finland
[2] IAI (2005) Efficient flow of information in the building process using IFC, Presentation Handout, BuildingSMART conference,
http://www.iai.no/2005_buildingSMART_oslo/June%201st/buildingSMART_live.PDF
[3] Ibrahim, M., Krawczyk, R. and Schipporeit, G. (2004) Two Approaches to BIM: A Comparative Study, In Education in Computer Aided Architectural Design in Europe, eCAADe 2004 Conference, Copenhagen, Denmark, pp. 610-616
[4] Cooper, R., Aouad, G., Lee, A., Wu, S., Fleming, A. and Kagioglou, M. (2004) Process Management in Design and Construction, Blackwell Publishing Ltd, Oxford
[5] An extrapolation of the meaning of the word ‘Model’ as appearing in the Oxford Dictionary of Philosophy: “A model for a set of sentences is an interpretation under which they are all true”
