buildingSMART Forums

IFC structure improvement

#1

Dear friends,

Do you think is it possible to change IFC structure a little bit?

The main one is:

Occurrences and types are further subdivided into six fundamental concepts: actors (“who”), controls (“why”), groups (“what”), products (“where”), processes (“when”), and resources (“how”).

I think in this way:

Who = Actor
Why = Control
What = System, Element (Product is part of Element), Part
Where = Project, Facility, FacilityPart, …, Building, BuildingPart, BuildingStory, Zone, Space
When = Processes
How = Resources

I think Product is not a good occurrence for where, where should refer to the project

IFC as four main modules, each optional
#2

We can think of it only we get something good outcome, but firstly why you think to change it, are you facing difficulties in using them.

#3

I think IFC especially after expanding it and adding IfcFacility and IfcFacilityPart, and IfcBuildingPart, etc which are familiar entities (adding Infra and other parts), needs some improvements to simplify and also strengthen its structure (ontology)

Something like what happened on SimModel

#4

IFC4.2 schema entities:

  • Over 45% of IFC schema is related to “Presentation and Representation” like Geometry/Topology, etc (326 entities)
  • Over 40% is related to Product mainly Element (268 entities)
  • There’re other small but effective parts like:
    • Relationship (60 entities)
    • Property (30 entities)
    • Quantity (11 entities)
    • Environment [anything related to location] (27 entities)
    • Resource (16 entities)
    • Process (8 entities plus 7 entities ( IfcSchedulingTime ))
    • Control (11 entities)
    • People [Actor] (5 entities)
    • etc

So, big parts are Product/Element and what I call Model which can be divided into “Presentation & Representation” which I’ve started to borrow it from Product Manufacturing Information (PMI)

#5

https://pre-scient.com/knowledge-center/geometric-modelling/faceted-modelling-and-nurbs.html

STEP AP 242 Ed2 and Ed3 (in the near future, end of 2020) are the best references to solve the current challenges related to geometry/topology