Beyond BIM – Building With Perfect Information

In this Viewpoint article, Ray Crotty, who is the founder of the UK-based C3 Systems and the author of the recent book “The Impact of Building Information Modelling,” contrasts drawings with models, asserting that with drawings, no matter how detailed you get, you can only say approximately what you mean or intend, whereas there is no such ambiguity in models. He also questions what would happen if the information used in the AEC industry were fully trustworthy – needing no checking – and were readily computable, as the BIM vision promises, and if the operation of the construction industry were based on the use of effectively perfect information.


14 thoughts on “Beyond BIM – Building With Perfect Information

  1. 1.Effective competition.

    Hmm, Maybe. It is certainly true that better coordinated and more error free information from the design team would leave less opportunity for the contractor to exploit.

    2.Manufactured buildings

    The point to remember about manufacturing is that the product is usually designed and built by one company, and that company invests heavily in the design of the product because they are banking on selling high volumes.

    The construction industry is fundamentally different in that the design and manufacturing teams are at odds to each other from the start.

    Buildings COULD be manufactured, and maybe they will – but not every building, after all Architects strive to design new and innovative buildings. There will never be time to design bespoke buildings in as much detail as aeroplanes.

    Houses – yes. High rise accommodation, maybe. Iconic buildings – no.

    And I really want to point out here that Revit is only used to model components too a 1:50 scale level of detail. We won’t be designing whole buildings this way unless we start using Catia.

  2. The analysis of the enormous inefficiencies created by the untrustworthy handoffs and approximate nature of the designs for the prototypes we build in commercial construction is right on. However, the author understates the complexity of the completed project and the interfaces and relationships among the various items/assemblies. You may be able, for which you can possibly model in enough detail to manufacture. Closer knowledge of a Gehry project would reveal how bumpy and not-foolproof of a process it is behind the curtains and the PR. The statement following the Gehry sentence: “The idea of ‘tolerance’ will disappear; individual objects will be manufactured with effectively perfect precision” reveals a lack of understanding of materials. Every manufactured item has tolerances and anyone designing anything physical in which materials must fit together must understand those tolerances. A lack of understanding of such tolerances is a common problem in today’s drawings. And questions of model ownership (and liabilities) often revolve around that issue: If material B sits on or attaches to material A, and there is a BIM for the building and a model for A and B, who is responsible for understanding the tolerances (both manufacturing and installation) so A fits well (and does not create a problem for C…)?

    This is not a software problem at this point, it’s a knowledge problem. In theory the rules can be embedded in software or a model, but we are much further away than the techno-optimists realize. My perspective includes both current work “in the trenches” trying to get portions of buildings built right the first time and membership in IAI in the 20th century.

    “IFC is too complex, too large, and too fragile to survive in the real world of live projects. Commercial IT companies are much more likely to produce a robust solution in this situation than committees of experts.” No, the problem here is that a live project is more complex in design than the author understands. Mr. Crotty believes that “designers need to deal in terms of generic components, not products; so a comprehensive, extensible catalog of construction components will be required.” I believe he grossly underestimates the complexity of constructing and maintaining any such catalog.

    Leo Schlosberg

  3. Another Revit “misinfomercial” posing as an objective paper by an industry expert.

    AECbytes, can we please have more of the likes of Lachmi Khemlani’s truly objective analysis rather than this kind of stealth advertising posing as fact!

  4. Your lack of appreciation of the 2D drawing astounds me. Skilled draftsmen — admittedly of which today there are very few — can and do produce excellent drawings that clearly define design and manufacturing intent. The basic premise of producing the drawing in the first place is to convey this information in a clear and unambiguous manner. Unfortunately this skill is being eroded as 2D draughting is no longer taught to the same level as it used to be.

    You also claim that the information presented is inherently untrustworthy…given that most 2D drawings for the last 10 years are directly derived from 3D models. By your logic, this would imply that the originating 3d models are also untrustworthy!

    I am a trained draftsman, migrated from the drawing board in 1985 to CAD, and worked with BIM projects for the last 12 years. I can honestly say your perception of the value of 2D drawings is fundamentally flawed.
    The idea to recognise Revit as the de facto industry standard is also questionable…Revit was not even the first product to adopt BIM!

    In section 4 of your article, you commonly refer to aspects of the design and build process as if they were a future consideration; most of what you describe is already an integral part of many projects and has been for a long time!

    However, I do recognise that there are still many companies in the UK that don’t understand BIM and the potential of intelligent modelling workflows. Before we can move to a full implementation of BIM as you have described, we must first address the real issues of awareness of the basic benefits of BIM, what it actually is, and its potential for the future, by presenting examples from companies already at the forefront of BIM integration in both the private and public sectors.

  5. Another idea:
    Remember the CAD-shootouts in the 90’s? AECbytes would be the perfect organisation to stage BIM-shootouts between the major players. That would provide some real world comparisons of the speed, efficiency, elegance, and scope/flexibility of the available BIM platforms.

  6. Having worked with collaborative process and technologies for 3 decades across a wide range of industries, I read Ray’s article with interest. I agree with much of what he says regarding manufactured/guaranteed buildings and changing roles. He also has a valid argument when highlighting the vagaries of competitive bidding as an output of imprecise proposals generated via traditional methods.

    However, I do believe that his opinions of interoperability are inherently flawed, both from a business and technical perspective. IFC is not a perfect solution – but neither is Revit, or any proprietary vendor led system, nor could they become a ‘de facto’ industry standard. However, IFC does work – its failings are often down to the unwillingness or inability of particular practitioners or vendors to make (or indeed want) it to work. IFC is also at the core of the government working groups BIM Strategy for level 3, so to ignore it now could mean massive rework of process in the future.

    IFC is owned and developed by the industry, for the industry and is continually improving. It’s about time that the industry realised this and gave their support rather than be wooed by the false hope of any commercial organisation providing a solution that aims to do anything other than promote its own interests.

  7. An excerpt from the article: “One of the biggest practical obstacles to effective diffusion of BIM in the industry is the lack of data exchange standards and associated protocols. The Industry Foundation Class (IFC) definitions being developed by BuildingSMART will be useful, but will probably be used mainly as an archive format. IFC is too complex, too large, and too fragile to survive in the real world of live projects. Commercial IT companies are much more likely to produce a robust solution in this situation than committees of experts.”

    Completely disagree with this. IFC is and should be the standard for information exchange between BIM platforms and not the Revit file format. The author clearly underestimates the requirement of an open standard file format for BIM collaboration. You can not achieve this by any proprietary file format such as Revit or Bentley or ArchiCAD. There has to be an open standard file format which talks to all BIM tools. Please see this blog post here talking about the same issue.

    Additionally, if you look around the world and look at some of the countires such as USA, Singapore, Japan, Germany, China etc., the government (as well as some non-govenement bodies such as AGC, army, universities, state governments, etc.) are mandating IFC as the output file format from BIM. Nordic countries are far ahead in this area too. While the AEC industry, other governments, NGOs, professionals, etc. are advocating for open standard file formats such as IFC, it is disappointing to hear this view coming from someone who is part of the UK Govenment BIM workgroup.

  8. “IFC is too complex, too large, and too fragile to survive in the real world of live projects.”

    It is true but that’s why BuildingSmart developed the IDM business process ?

    “So why not accept the facts as they are and recognise Revit (for now) as a de facto industry standard?”

    This will most likely never happen ! Fortunately.

  9. Great idea, let’s rely in Autodesk for the future of the construction industry!

    We’ll use Revit for all our needs, and if Revit can’t do something … oops. Guess we’ll need to ask them to write some more software we can buy off them to fill this need, or just accept that this part of the industry isn’t going to benefit from BIM.

    Isn’t it amazingly obvious that one package (or even a suite) can’t cover all our needs? And opening up old versions of a proprietary file format so we can all work with a superseded file format is almost laughable.

  10. About IFCs supposedly problematic complexity: if complexity means rigourous organization, rules, and standards, then yes, IFCs are complex. As a building is, no more no less.

    Seems the author want to go back in the 80’s when DWG won the race, but a race with no competitor. Let’s try again! Bad news, there are competitors. And Autodesk is at 15/1.

  11. IFC, in the end, is a way to organize / categorize information. Revit is a way to create information and it’s stored in a format that is complimentary to that creation process (which might not be complimentary to how other software creates this information).

    This leads us to a point… it’s not important how the information is created and stored as long as it can be leveraged. Thus the way Program “A”‘s database is stored/organized needs to be able to be reorganized into something Program “B” can understand. IFC is the proposed solution to that. Is it the best solution, long term? I don’t know. But a “Revit” format as the defacto standard doesn’t make sense as it’s really NOT as good as IFC (which is more flexible based on the types of software programs that need to leverage it).

  12. So I suppose the work that 99 international vendors/developers (this includes 147 apps and utilities and counting)
    have put into IFC is just a silly side project to creating true interoperability for data in the AEC industry?

    I have found that anyone who hacks on IFC generally doesn’t have enough knowledge, comprehension, or experience with it to see what it is and the capabilities of an IFC-based, or at least supplemented, workflow.

    When people say IFC doesn’t work, it is usually because they are using products with poor implementation. The poor implementations are not because IFC is too hard, complex, or inflexible, it is from a lack of sufficient resource devoted to making it work.

    The new IFC2x3 Certification 2.0 program , by buildingSMART, is going a long way to prove IFC capability and also improve implementation quality for vendors and end users.

    But IFC-based/supplemented workflows also require end users to really think about how their workflows (and even modeling), change when they understand sharing data. Why should sharing building data – geometry, semantics, and attributes – be any different than the data we currently share today in so many other fields, via the internet, using open standards like HTTP and HTML?

    In the software industry today, there are many different technical means to connect proprietary systems using open standards. Unfortunately, the AEC industry is one of the last ones (if not the last) to step up and apply this knowledge effectively, across the board, to make itself more efficient, more productive. I think in large part, this is due to the ignorance of the end users and the lack of pressure they put upon the industry software vendors to get it done, done right, right now.

  13. The author has struck a raw nerve in suggesting the use of Revit as an industry standard rather than IFCs. I agree with others who have responded that this suggestion is not sound and we that we need to make IFCs the agreed upon goal for open interoperability. I also do not think that his suggestion of using the final BIM model for competitive bids is a good idea because it implies that the contractor and major subs did not take part in the development of the model which is antithetical to the idea of using BIM to support collaborative design and construction (with owner participation).

    However, his other suggestions have some merit, particularly the goal of incorporating more off-site fabricated components into the building and using an accurate model to support this goal. Trying to fabricate the entire building off-site presents many problems, but large modules and components are certainly possible so long as they are incorporated into the design from the start AND that the designers work with the fabricators and contractors from the start. This cannot be tacked on after a schematic design is completed.

    We should take from this forward-looking article the ideas that have real merit to advance current practice.

Comments are closed.