Saturday, September 5, 2015

Static IFC, or STEP + dynamic XML?

By Pedro Martín
-
The more I think about IFC, the more questions emerge.
.
I have spent some time thinking about using IFC with more specialized disciplines like industrial engineering or even something so far as aeronautical maintenance. Sounds fantastic; in fact sometimes sounds like a complete fantasy to me. And it probably is, but it have conducted me to a concept that I think it has to be taken into consideration when we think about a BIM standard for the building industry.

3D + Static Data Scheme
IFC standard file (IFC-SPF) follows the definition of the STEP-file defined in ISO 10303-21. STEP format (*.STEP, *.STP) is a generic structure used to define 3D models, that allows the definition of objects and its properties. Since it is generic it could be used for any use, but as long as it is, each specialized STEP use has to define a set or library of objects and properties ('STEP-lib') adapted to its specific discipline.
.
In this line, IFC is an STEP-File that uses a set of definitions focused on building industry. There are other examples of these specific uses; ISO 15926 format is another STEP specialized in electrical, piping, rotating equipment, etc. at oil and gas production facilities. Most of CAD systems export its proprietary formats to STEP.
.
Since all specific formats based in STEP could use generic definitions too, any specific format could be used for any other discipline. IFC has specific definitions (IfcWall, IfcBeam, etc.) but it also has a generic IfcProxy that can be used to host any entity that it is not natively defined in its specifications. So if we plan to connect that file with a different discipline application, we could add a description that would be identified and used at the other application via a mapping table. That way, IFC could be used for anything, Sky is the limit.
.
But the sky is too high - it is not efficient to use just IFC specifications to design an airplane since it would be full of undefined IfcProxies. The right way has to be to use a specific format based on STEP with a library of appropriate definitions. As I have said, ISO 15926 is the IFC for Oil & Gas, and I am sure there is some kind of STEP-based standard for aeronautics.
.
Rosseta Stone
So if we are searching for a kind of 'Rosetta Stone', it should be something like a STEP - based mapping table between discipline-specific definitions. As Rosetta, it probably would be used only on very specific situations -Memphis Decree doesn´t happen every day. But it has to be defined just in case it is needed.
.
I have to add something, though.
.
Building object properties are mostly static, so the IFC alike file that may be used for O/FM phase could be one with object instances and properties that will be overwritten sometimes, based in a building maintenance schedule.
.
But those other disciplines like oil & gas, aeronautic or even military industries have to deal with an enormous amount dynamic properties. Such a great amount that Big Data Management enters into place.
.
Currently, the 3D model part is the big one in BIM models. I am sure that 3D LOD will increase, flanges will have bolts modeled and so on, but I think that dLOI data part will be the real deal, especially for O/FM. In the future, I think that we will have to manage a small set of 3D entities linked with a big dynamic set of properties, not as big as the production industries but not small and static as it is now.
.
So I think that probably the most efficient way is to separate 3D and associated data in two (or more) files linked by an ID (as those STEP currently has, that '#3167'). This way we could manage the differences of size and dynamicity that a STEP file info has, letting them grow at that different speed they seem to demand. The 3D part could remain as STP, data as XML.
.
I'm not a programmer, so I´m not completely sure if this can be done or, if it is, it is operative. I have to talk with a specialist - this will be my next step.
.

No comments:

Post a Comment