October 23, 2013

IFC for content, follow up

Thanks for the warm welcome!
Not counting the inaugural post, I'd say the first "real" MdR Advies blog made quite an entrance... So to start: thanks to all of you who took the time and effort to respond to this post! It was very insightful, and I learned a lot!
Special thanks to Jon Mirtschin from GeometryGym though. This guy never fails to amaze me with his profound knowledge of what IFC is, encompasses and what it can become. His blog in response to mine gave great insight to current IFC possibilities and limitations (mainly related to implementation) and what efforts software vendors, even the most IFC-minded, have to make to get this format to work. As always it was an inspiring read that brought me a lot of new ideas, and lead me to abandon some old ones for being wrong, short-sighted or just simply outdated.
Oh btw: I'm wrong all the time. I know this. I just don't know when. So if there is anything that I put on this blog that you're sure is complete and utter bullshit, let me know. Frankly, that's partly why the internet as a whole is such a great thing. There's always somebody out there to learn from. My first blog had a staggering 600+ pageviews until now. But *only* a handful responses. Surely there must have been others that were thinking "what the hell is this guy blabbing about?! I know for a fact it's flatout wrong". Please do share your insights, I want to know. How else am I going to learn, right? I might not always agree with you, but I will always listen to what you have to say.
So to wrap this up: thanks again for the warm welcome in the blogosphere!
Why the follow-up?
Well, after the post there was some Dutch twitter exchange about it. Then I got a call from another Dutch guy. We talked about IFC, OpenBIM and my view on content libraries for BIM. I explained my view that creating those, as if we are back in the 1980's when we created the habit of dwg-libraries is insane for BIM. Basically we have a database that holds all product specs at a manufacturer. They then create pdf fact sheets and dwg drawings. And then they pay someone to recreate what basically is another database from those specs and drawings. Multiple times, for different BIM formats...
Then he asked me a question: "What are you going to do about it? Is this just another opinion, or will you be acting on it?"
My initial response was quite frankly "what the hell do you expect me to do? I'm just one simple loudmouth from a tiny country with no connections or influence whatsoever". Then, as I was driving home (well, more as I was crawling home due to some dumbass that parked his car in the highway railing), I started reviewing my day. Cause it had been an interesting day.
That morning I gave a presentation to the joint BIM managers of one of the largest Dutch construction firms about the Dutch Revit Standards. And they were enthusiastic above and beyond expectation. We had a wonderful discussion about the current status and future developments. Things that the Dutch Revit User Group was aiming to accomplish, and why. Basically the kind of meeting that makes you want to do a victory dance on your way out.
The afternoon was filled with a client meeting where we discussed possibilities of upgrading their 25 year old CAD-based database structure to plan and design major warehouses and connecting and integrating that with Revit. Opening up doors to dynamic model checking on layout-rules and such. Again, a meeting that usually leaves me tossing and turning at night, totally pumped with adrenaline and anxcious to get out of bed and start working.
So, "interesting" quite frankly is a gross understatement of my day. And as I was crawling along, slowly making my way home it hit me. If you told me June 2012, when I said yes to this crazy idea of singlehandedly creating a nationwide Revit standard, that in no more then 18 months I would be talking to one of the largest construction companies in Holland about how they could push this initiative further most effectively, I would have had you committed to a psychiatric hospital. But here we are...
So who am I to state that size matters? That just having a good idea and act upon it never pays off? That one guy cannot pull something off so big and groundbreaking?
Let's see where we stand Januari 1st, 2015. That's the mark.
I want to have at least one manufacturer that contributed to a good cause. If you know (or even better, work with) a manufacturer that is interested on getting on the BIM train, or already is on that train but not quite happy with the way things are going right now, drop me a line.
I'm not proposing to create a BIM library for you. I mean I would (even the most idealistic have to eat), but that's not the challenge. My proposal is to create a new way of handling manufacturer content:
I'm proposing to create a database structure that can be hosted online somewhere.
For that structure we need some sort of definition of geometric and non-geometric properties. Don't tell me that's not doable, each manufacturer with a CAM production facility has it.
This definition needs to be open source so that any manufacturer willing can translate his own products into this format.
We then create plugins for participating BIM software tools that can query the database, find relevant product definitions and take them to automatic object creation in the BIM software based on some simple rulesets and mapping tables.
Is this going to work? Let me put it this way: I know this can be done between Revit and several types of databases. I've seen it work. I don't know if it will work with other software. But I would guess it does.
From there, it's all just a matter of finding the right people willing to spend some time doing it.
Why would we? Welk for one thing I'm sick and tired of all these partial libraries scattered around the internet.
And me having to redo them on every freaking project that has a different Revit standard.
I hate having 30+ libraries from different manufacturers, each structured differently.
I hate spending hours trying to find that one freaking component with some specific properties that I know is out there somewhere.
I hate having to upgrade all that shit every release only to find out it doesn't work as it's supposed, needing to go online 30+ times and re-download all those libraries and having to go through all the painstaking adjustments again.
I hate it when I get my balls busted by a General Contractor because a model in my project holds components that aren't even being sold anymore.
And when I was a manufacturer, I would hate to have to create a library of products for 3 or 4 different software formats. In god knows how many countries. And still get continuous bitching and moaning from people claiming that it isn't good enough. Or just flatout wrong.
If you, as a manufacturer, thought that creating and maintaining a CAD library was expensive, welcome to the world of BIM. You just quadrupled your budget. At the very least. On each software format you want to provide.
And no, Revit is not the new Autocad. Just rvt's isn't the same as just dwg's. It won't cut it.
So, at the chance of monumentally overplaying my hand here: who's up for a challenge???


  1. Let me look at this from the perspective of a revit user, more specifically a revit mep user. When designing, I definitely need generic libraries created in native revit format to unleash all intrinsic power of the software. On the other hand, when moving onwards to a model for facility management I can see the benefits of an automatically generated component for manufacturer-specific content. This does not mean that I would not like to have native revit components for manufacturer-specific content! I'll keep promoting this but do understand that not all manufacturers can support all formats. Due to the highly specified software the number of formats will surely diminish in the future. Standardisation still remains an issue though...

    1. But I would not want anything else then native content. I just want it to be created automatically from an external data source. But NO imported 3Ddwg's or something. Native content.

      Having said that, I do believe that it's not suitable to use manufacturer content during the design process. We need generic content in this stage. However, when we have a manufacturers database, a generic database would be the same. Just without manufacturer specific content...

      The main idea of a database driven workflow is that manufacturers don't need to support all formats. In this workflow, the responsibility of creating the actual objects in all different formats no longer lies with the individual manufacturers, but with a centralised plugin creator (being either an open source initiative, software vendor or third party developer).

  2. I will give you a response live, when you are in LA in a few weeks but great job, next step: close the deal(s)

  3. I assume that technically, any relevant BIM software could be improved so it supports the approach you are proposing.

    To me it seems feasible if the manufacturer can enter their products just once and update them in real-time. I would however be careful about the practical implications of such an approach: as architects, we often design using generic components and products and leave the specification into actual manufactured content to the contractor.

    How do you see the workflow evolving? Will architects need to shift generic objects to specific ones in their design model? Would this be automated? Or would the contractors propose changes that can reflect back into the design model? How to manage changes? How to deal with deprecated content?

    1. Hi Stefan,

      Basically you provided the answer yourself: this isn't meant for generic components but for manufacturer specific components... So I don't see the problem there.

      Nonetheless, it would be great if manufacturers with similar product lines start talking amongst each other and provide (rules for) generic content that can be later swapped for manufacturer specific content.

      Looking at it from this angle I also don't see the need for changing workflows. It's basically the same as we do now (replace generic components for manufacturer specific content), just not manually, based on old and possibly incorrect libraries but with a live connection to manufacturer provided data.

      As for deprecated content: this would be the beauty of database driven content. It would be fairly easy to build a checking system that looks for objects in the model which are no longer manufactured.