OSArch Community

BlenderBim IFC Certification?

  1. S

    Hello all,

    I'm doing a competition that emphasizes on BIM models. They are a bit mentally tight to mainstream software like Revit.

    They asked me if the software I'm planning to use is on this list

    https://technical.buildingsmart.org/services/certification/ifc-certification-participants/

    BlenderBim seems not to be here!

    But BlenderBim and FreeCad are in this list:

    https://technical.buildingsmart.org/resources/software-implementations/

    I've just played a bit with BlenderBim, I want to say congratulation to all developers because the result is impressive! (I'm an architect and a developer to;) )

    Question:

    Is BlenderBim is IFC2x3 Certified anyhow? maybe @Moult you will be able to instantly answer that question.

    Thank you so much, I would love to use something else than Revit!

    Yann

  2. B
  3. S

    Yes @bernd it makes sense. Thank you so much

  4. D

    I hope that when relevant free software projects are ready we can apply for funding or do something to get them certified. That would certainly raise some eyebrows in the industry.

  5. R

    Who cares about the bSI certifications?

    IFC is going to disrupt, I will destroy it soon :))

    Oh, I forget, it's a standard!!! Really?

  6. S

    @ReD_CoDE said:

    Who cares about the bSI certifications?

    Contractors care about IFC and Revit because it's a feeling of security for them even if it's not true.

    I need to prove to them that BlenderBim is the right tool for BIM oriented architectural model despite their believes. But I need some official stuff ;)

  7. R

    It's up to Dion and other friends, but personally don't care about some companies/contractors who just like show off, but even don't have average knowledge needed in their profession

  8. M

    @superseed77 I share your concern. However, it is currently cost-prohibitive to certify. As @bernd shows in his price list links, it actually takes 11,000 Euros just to get access to the certification "platform". If we want to certify import, it costs 25,000 Euros to begin, and then an additional 3,000 Euros for each time you run the test. That's just import. Export is 27,000 Euros. There are some bulk discounts, but basically you're looking at 50,000 Euros conservatively to have a few tries to crack things as nothing is perfect. Oh, and that's just for a single MVD. Multiply the cost by each MVD.

    BuildingSMART can charge these numbers because they primarily address the big vendors, leaving the smaller ones out. Even then, I've been waiting for IFC4 certification for Revit for a while now, and it hasn't happened.

    I've also tried talking to buildingSMART about offering discounts or options for free software, with no success. I have also proposed a leaner certification process with automated unit tests for "self-certification". No luck.

    For what its worth, you have my word that the IfcOpenShell project and the BlenderBIM Add-on project have a huge focus on valid IFC processing. My own experience is that free software actually does a better job in all aspects than commercial software in IFC: importing, exporting, minimising data loss, preserving fidelity of metadata and relationships when authoring natively, and supporting broader, even more esoteric aspects of the IFC schema.

    Contractually, where I have influence, I remove any mention in the contract about IFC certification. Instead, I reword it to talk about exactly what data in the IFC we will check. Any software capable of producing that data is allowed. If it is not capable, regardless of official certification, then it is not allowed - after all, what's important is the output, not the certificates.

    I'm also not sure how much faith I have in the certification process. For example, they certify "colours in IFC", but that has been broken for quite some time in certain certified software. I guess if I had 50,000 Euros to invest right now, I wouldn't spend it on certification. I'd invest it in software development. It'll go much further.

    More interesting links: https://blenderbim.org/blenderbim-vs-revit.html

  9. S

    Hi @Moult! Yes, the cost is totally prohibitive. I totally agree with you. And yes @ReD_CoDE many of them don't even understand the purpose of BIM in architecture, they only need to hear this word next to Revit and that are happy!

    Thanks a lot, guys. This project is awesome. You have understood the practice.

    I Have a big project with three buildings and many floors at mid-level. In Revit, all those things are a pain in the ass and you have to use workarounds to make it work.

    In two or three years this tool will make the architect gig very loveable again!

    One question as a newcomer: Do you think it's safe or not to use BlenderBim and build the model in order to make sections for the floor plans and sections? Is your tool mature enough to make a full preliminary design file with it?

    Thank you

    PS: I 've been able to 'sell' your tool as a Revit-compatible tool ;)

  10. M

    @superseed77 personally, I do not yet see BlenderBIM as safe to use to deliver drawings. The drawing generation features are quite rudimentary. I have used it myself to deliver a building, but that's because I know all of the pitfalls and can fix them ... there's nothing quite more stressful than having to get out drawings the next day, and you haven't even written the code that let's you create spot levels! It's certainly motivating, though.

    I do encourage you to attempt to use it nonetheless, and once it falls short or you aren't sure how to do something, post here and that feedback you provide will be incredibly valuable in helping us fix and improve it. This feedback loop of users and developers working together is vital to make things ship shape sooner rather than later :)

  11. B

    @superseed77 said:

    One question as a newcomer: Do you think it's safe or not to use BlenderBim and build the model in order to make sections for the floor plans and sections? Is your tool mature enough to make a full preliminary design file with it?

    Give FreeCAD a try. It is suited for this. But be aware most Architects who use FreeCAD for design need to finish the drawings exported from FreeCAD with some 2D drawing software like LibreCAD or Inkscape.

  12. B

    BTW: even if a software is certified it does not mean it exports standard ifc. It only means it has passed the certification test. These test for sure do not cover all user problems. If they would we would have much much less problems while importing ifc from proprietary vendors.

  13. M

    Whitout reading (all) of the above in detail...

    My reaction to IFC certification is this: it is poor and standard for not much!

    BSI community to me seems like 'closed community' not much openness there

    Frankly, these IFC stickers on software boxes has little value. Not these amounts.

    Let user experiences and forums like this be users's guide.

    Go Open Community!

  14. S

    @Moult Thank you for your answer. I know how stressful it could trying to do and fix with a deadline ticking clock! I would give a try and if I have enough time I'll go for testing drawing generation. I will tell you how is going.

    @bend thank you for the Freecad advice :)

  15. B

    Why not write a test suite that covers all the important things that actually need to be done with IFC for it to be a first class file format (we need this for our own purposes anyway, there is already a start of this with the roundtrip tests), bundle it with a validator, draw an official looking stamp, and call it 'Future IFC' certification?

  16. R

    @brunopostle there's a long time I suggested this to Dion, and also some bSI guys, but there're a lot of reasons they don't want to have a strict certification framework (even Dion, as part of IfcOpenShell movement, which is part of bSI movements too)

    IFC, even IFC4x3, is too small to cover things today all we want, and they're too slow and too close too, for this reason, I decided to develop our own schema/file format

    And many have started this approach too, from Google, to Spackle, to others

  17. M

    @bernd and @magicalcloud_75 are correct - just because a software is certified doesn't mean it exports standard IFCs. The certification process doesn't handle real-world usecases. I've tried to address it, which I'll explain below.

    @brunopostle I'm not sure what @ReD_CoDE is claiming about me not wanting strict certification - I've certainly mentioned quite a few times to buildingSMART that a test suite is needed - it's the textbook answer any software developer would give. buildingSMART has ignored it, presumably because they get a lot of money from the certification process.

    Creating a test suite is quite a bit of work, but I've already made quite a bit of progress - would you like to contribute? Here's what I've done so far:

    • Create BIMTester to allow for automatically testing and writing test specifications for IFC files. This is the validator you speak of. It uses existing unit testing standards and spits out JUnit XML.

    • Create MicroMVDs as an alternative to MVDs, which are used in certification. This allows you to test smaller aspects of data manipulation than 1,000s of entities in the certification MVDs. These are the "test suites".

    • "Certified" (i.e. tested myself) the BlenderBIM Add-on to these suites:

    • Project setup

    • Geolocation

    • Model federation

    • Element classes

    • Geocoding

    • Naming

    • COBie (work in progress)

    • Created a test server, currently very incomplete, with test file repos available here, and here and here.

    • Those test suites do not cover parametric geometry. For that, @theoryshaw and @yorik have been leading the way in their tests, although it is not automated. I am catching up to their progress, and when I have done so, I will automate a small portion of their tests. Note that their work is extremely valuable, since it tests the user experience in round-tripping, which cannot be automated fully.

    Maybe this is worthwhile giving a presentation on to explain the approach and work done so far?

  18. B

    @Moult this sounds awesome, we should help. Part of the problem for me is that there are whole branches of this technology that are a complete mystery. For instance I have no experience of MVD, or have ever encountered its use in the wild.

    My observation, and it may seem a bit glib, is that since existing official certification is of no interest, any tests that we have that are useful across tools should be bundled together and given a name and logo (BIM++ Benchmark?). This will suffice for most QA purposes, where all that is needed is a genuine documented procedure that is demonstrably suitable for the task.

  19. M

    @brunopostle yeah, there are issues with MVD, one of which is how foreign it is to everyone. You may be interested in Compliance checking on building models with the Gherkin language and Continuous Integration - a paper co-authored with me and @aothms , which I've now added to the Wiki AEC academic paper directory.

    Yes, the process should be properly packaged and branded. There are so many things to do :) Maybe I can present the work during an OSArch meetup, and we can discuss where to take it.

  20. R

    @Moult the problem is that anytime someone shares an idea with you, you try to implement it yourself, based on your view

    I shared with you and also bSI friends this one, please check carefully:

    Step by Step test suits, like what's common on PLM area like CAx-if test suits

    Write a complete test that follows some steps and each step checks a specific thing, for instance, one checks header implementation, another one colors, another one materials, another one meshes, another one solids, another one Breps, another ... then share it with all and say for instance BlenderBIM passes these tests, but Revit passes that tests

    In this way soon bSI has to accept a better approach than the existing one

  21. M

    @ReD_CoDE the tests are based on industry standard JUnit XML. None of it is my own invention. Even the contents of the test suites are based on buildingSMART IDMs.

    If you read the test suites, you will find that they are step by step, each checking a specific thing. This is standard practice in unit testing, exactly as you've described. I'm not sure what your concern is.

  22. B

    @Moult I worry about you spreading yourself too thin, you can't do everything, but yes I think we all need an introduction to BIM testing.

    I wonder how continuous integration testing fits with the existing AEC industry (where everybody is intentionally hiding information, being deliberately vague, and not committing until the last possible moment). Perhaps this is the kick the industry needs, but the proposed change in culture is generational.

  23. R

    @Moult you came back to IDM, do you remember I constantly mentioned the main problem is IDM and bSI works on IDM, so let's build an IDM Toolkit which is better than IDM Toolkit/Configurator bSI develops?

    IDM Toolkit is based on BPMN (http://big.yonsei.ac.kr/xppm)

    IfcXtreme (https://www.linkedin.com/pulse/ifcxtreme-ehsan-azari) shares something vitally important for those who can understand what I did share?

    I shared the solution, the idea that will completely change the industry, but few can or will realize what I shared?

  24. D

    I shared the solution, the idea that will completely change the industry, but few can or will realize what I shared?

    Remember @ReD_CoDE we talked about humility a few months ago? If your solution is the best one then it will gain support. Telling everyone else they're doing it wrong and you're the only one who understand the problem, just annoys people (me at least) and gets in the way of the insights you're trying to share.

  25. B
  1. Page 1
  2. 2

Login or Register to reply.