Tuesday, January 11, 2005

BIM - What is it, why do I care, and how do I do it?

"BIM" or Building Information Model(ing) is perhaps one of the most overused and misunderstood terms in the architectural industry today, on several levels. The term "BIM" is sometimes used anytime someone generates a 3D model of a building, or a part of a building, whether that model has any non-graphical information or not, or regardless of the level of detail that the model may be done at. Part of the problem is that Building Information Modeling is an evolving concept; one that will continue to change as the capabilities of technology and our own ability to manipulate that technology improves. It's hard to tie it down with a simple definition. Yet more and more people are jumping on the "BIM Bandwagon". So rather than try to define Building Information Modeling, I'll try instead to lay out some of the issues and benefits to be gained by pursuing the holy grail.

First, emphasis needs to be placed on the "I" in "BIM" - "Information". That information can be either graphical or non-graphical, either contained directly in the building model or accessible from the building model through linked data that is stored elsewhere. If you really think about it, in some ways, at a basic level "BIM" doesn't necessarily require that the geometry that describes the building be a 3D model at all. A data model is just as valid a model as a geometric model. Be that as it may, when we think of a Building Information Model, a 3D geometric model of the building is at least part of what comes to mind.

So what kind of information can we put in a model today, and why would we want to?

1) Geometry: Obviously the more we can get into a 3D model, the more we can communicate our ideas and design intent graphically. Additionally, most BIM technologies today allow us to generate elevations and sections directly from the 3D model, and in most cases, allow for those 2D representations to be directly linked to the model. When a change is made to the graphical model, the views generated from that model can be updated with no further editing, cutting down on errors. Additionally, creating a 3D building model allows us to identify design issues at a much earlier stage in a project, resolving them before they become potentially expensive change orders.

2) Non-graphical information: With today's object-oriented technology, the graphical model components can themselves contain non-graphical information that further define their function and structure. For example, a wall in Architectural Desktop can have any number of non-graphical "properties" assigned to it that contain information about fire rating, framing parameters, estimating information, etc., that would not be obvious from simply viewing the model or a 2-dimensional view. This information can be displayed in a schedule or exported to a 3rd party or complimentary application for downstream use by another discipline.

3) Linked information: Information related to the model, but not contained directly in the model, can be linked through hyperlinks. Examples are manufacturer's cut sheets, Gandt charts for construction scheduling and specifications just to name a few.

Why would I want to include that information? Who would use it?
While many architects are a bit hesitant to put too much information out there due to liability issues, it is my belief that the adversarial nature of the construction industry in the United States will inevitably change. The costs of building these days are becoming more and more inflated due to legal issues and malpractice insurance. At some point, the industry will be forced to change, or nothing will be able to be built. In fact, I think we're already seeing the beginnings of that change, with more and more design-build firms coming into existence, and more owners who realize where the real costs are and who are looking for ways of eliminating them.

So, liability issues aside, putting more information into your model has the potential to benefit your consultants and collaborators during the design and construction processes, and can further benefit your customer, the building owner or manager, once the project is complete. The term "Building Lifecycle Management" goes hand-in-hand with "Building Information Modeling". Are most of your consultants, collaberators and customers capable of using the information to it's fullest extent today? Probably not, but they are becoming more and more so. For example, a structural framework can be exported from ADT as a 3D "stick" model with connection nodes included. This stick model can be processed with structual analysis software for correct sizing of structural members. With the powerful API (Application Programming Interface) of ADT, it's not too much of a leap to visualize an application that then puts the results of that analysis back into the graphical ADT model as correctly sized structural members. There are already several third party structural products for AutoCAD that can communicate with Architectural Desktop. Additionally, estimators can utilize the combined non-graphical and graphical information to generate cost estimates faster and more accurately, as well as react to design changes in a more timely fashion. The list goes on. Consider facility-management software that can plug into the finished "as-built" building model data to assist the owner in the long-term operations of the building.

And consider further the prospect that the building owner, recognizing the value of this information, may be willing to pay for it, over and above the standard architectural services fees.

What are some of the barriers or issues that I need to consider if I want to start moving toward Building Information Modeling?

First, realize that you won't simply start "BIMing" tomorrow. It's an evolutionary process. And one that you may be ongoing for the foreseeable future. Again, "BIM" is still an evolving concept - who knows where it will lead, or if it will ever stop changing? Also realize that most, if not all of the issues surrounding and barriers to the adoption of Building Information Modeling are technology independent.

Processes will most likely have to be studied and, in most cases, redesigned to fully realize the potential of Building Information Modeling. Most of today's firms are still relying on traditional 2D design and documentation processes. Even though we're using powerful CAD technologies to get the work out, the paradigm is not much changed from the days of manual drafting with T-Square and Triangle. Building Information Modeling will change these processes. Design problems will have to be resolved at an earlier stage. It will be harder to define the lines between schematic design, design development and construction documents. In fact, there may be no firm line between design development and construction documents - the two processes will become more and more concurrent as the evolution progresses!

Changed processes may very well lead to staffing changes. In the past a Project Architect, with extensive knowledge of building construction could oversee a "pool" of drafting technicians and interns with less knowledge of how a building goes together. With Building Information Modeling, anybody who is contributing to the model will need to have a good understanding of how that building really fits together. The need for 2D drafters will become more and more a thing of the past. At the same time, architects who in the past may have shunned CAD and similar technologies will have to become more comfortable working in a digital world. Additionally, you may find that your current CAD support staff is inadequate to handle the expanded needs of an office involved in model-based design.

Culture is, without a doubt, the biggest issue that is faced by most firms attempting to change their processes and make the move toward Model-Based Design (that would be the first step to Building Information Modeling). Users are dealing with change, sometimes large-scale change, to the ways they are used to working and dealing with technology. The normal human reaction to change is to resist, and the more change we're faced with, the more we try to resist. There are no panaceas for dealing with these cultural issues besides patience, perserverance, and most importantly, commitment by upper management. Without firm and unwaivering upper management commitment, the endeavor will be doomed to failure.

Level of detail is another issue that you'll be dealing with constantly, no matter how far along the BIM trail you are. Some questions you should ask yourself whenever you are adding geometry to a model are:
1) Should I model this in 3D? What is the benefit of modeling a component weighed against the effort required?
2) How much detail should I put in the 3D model? This is a question you should consider frequently. As technology (both hardware and software) improves, and your ability to manipulate the model improves, the amount of detail that you'll put in a model will grow. As a general rule, I've found that anything that would only show at a scale of around 1" or 1 1/2" inch equals a foot is something that you probably would not bother to put in a 3D model. There is still a need for 2D details, and this is where these types of components would be shown.
3) Should my details be model-based? I like to categorize details into 3 main groups. Model-based details would be sections and elevations that are linked directly to the 3D model and that update with the model. Hybrid details are details that might have their major components generated from the model (such as a wall section), with smaller components (framing, fasteners, vapor barriers, etc.) drawn on top of the model-based elements. 2D details are created completely from 2D geometry and have no real link to the model. Things that you detail at a scale of around 3"=1'-0", for example, will frequently be purely 2D in nature.

Technological competency on the part of most users will have to increase greatly. Users will need to be able to deal with the model and the model output at a level to which they are most likely unaccustomed. The term "work smarter, not harder" could apply here, and that means that users will need to become more knowledgeable about and more comfortable with the technology that they're using.

Content is another huge issue facing anyone who is involved with model-based design. The task of developing symbols - especially symbols that have a plan, 3D and elevation representations will be a never-ending one, and one that you will have to budget resources for on an ongoing basis.

With an understanding of these issues and barriers, how do I go about moving to Building Information Modeling?

Start slowly. First, make sure you have a good set of standards in place, and that all participants are educated in those standards and abide by them. If you're using Architectural Desktop and are practicing in the United States, the Imperial templates come with layer standards and a display system that is compliant with the National CAD Standard version 2.0, and have pen style tables to match. Additionally, many of the annotation symbols are compliant with NCS. You can either adopt that standard and use the templates and other content that is provided, or you can take the time to develop your own and set up the templates and content to suit your standards.

Contract with a knowledgeable source for training and ongoing consulting. It's doubtful that you have all the expertise in house to get your users up to the level of use of the software that is necessary for model-based design, or you probably would already be doing it. You may very well also benefit from an objective third party analysis of your current processes and suggestions for change that are free of bias towards "established traditions". The term "If it ain't broke don't fix it" may need to be changed to "If it ain't broke, break it."

Phase your evolution. Begin with gaining more productivity with 2D aspects of your software and utilizing more of the "I" in "BIM". For example, if you're using Architectural Desktop you might start using the scheduling capabilities and some of the drawing coordination features (which would entail diving into the world of the Project Navigator).

Once you have gained familiarity, productivity and comfort with those tools, then you can begin to work on basic three dimensional models and try to carry them as far into the construction documents phase as possible, utilizing the automated sections and elevations. These initial models will probably be at a lower level of detail, but again, as you develop your organization's skillsets, as the technology improves, and as you develop your content libraries, you'll find your models increasing in detail as well as longevity into the design/document/build/manage processes.

Building Information Modeling and Model-Based Design are potentially the biggest changes to hit the AEC industry since the advent of computer based design, and should not be undertaken lightly. As a collegue of mine says frequently, "If it was easy, everyone would be doing it!" That said, more and more people are moving to BIM, and are beginning to realize the very real benefits that it offers. Will you join the movement now, while it's still in it's infancy? Or will you wait until it becomes more mature and find yourself playing catch-up with your peers?

For further reading on this subject, I offer the following resources:

BAA Heathrow Airport Terminal 5 Project - of special note here - the owner, BAA, absolved it's contractors from all risk, allowing them to focus on the teamwork necessary for a successful true BIM project.
Navisworks Case Study - BAA and Laing Industries - more on the Heathrow project with additional BIM links.
The Lean Construction Institute - Non-profit organization dedicated to disseminating information related to project management with lessons learned from the manufacturing industry.
Graphisoft's Virtual Construction Modeler - Although not an Autodesk product, this is of note. Read about an application geared for the construction industry to model a building at a construction level of detail to assist contractors and suppliers with estimating, phasing, ordering and staging of materials and trades. This is not a competitor for ADT or any other Autodesk product currently on the market, but has huge potential to be complimentary as part of a total BIM solution!
Building Information Modeling Debate - Debate between Phil Bernstein of Autodesk BSD and Keith Bentley - co-founder of Bentley Systems. Hosted on Jerry Laiserin's web site.


Blogger Wes Macaulay said...

This is a really good article. Every principal should read this! Whether it's ArchiCAD, AllPlan, Triforma, ADT or Revit, the issues are much the same. Nice work, Matt.

12:02 AM  
Blogger Jack Talsky said...

Thank you for taking the time to explain a term that I see used all the time, but whose definition seems to be vary with each user.

3:29 PM  
Blogger Justin Rigg said...

Thank you greatly Matt. I have forwarded this throughout my office and United States offices. In an effort to back up my demands for better 3D modeling and non-graphical information content.

10:38 AM  
Blogger Justin Rigg said...

Thank you sincerely Matt. I have forwarded this article throughout my office and our US offices in an effort to reinforce my demands for better 3D modeling and non-graphical information.

10:40 AM  
Blogger Dave Baldacchino said...

Matt, almost three and a half years later and this text is still very relevant! Of course the talk is more about Revit, but the way you worded this article, it is very software-independent. I keep sending this to people that want to understand what BIM is, and then explain where we're headed, with the advent of more software that takes advantage of the proverbial "I", such as automated code checking, specification generation, etc. Awesome job!

8:15 AM  
Blogger Virtual Builder said...

Matt, our small architecture firm in MD has been using BIM since 2003. While explaining the service to clients, we have hit some roadblocks trying to make people fully understand the importance of the information held in these models. You've made everything very clear and easy to understand, which will surely help us in the marketing of this innovative system for the future!

9:59 AM  
Blogger Unknown said...

Thanks for this informative article.
Just one question: what impact will this have on the HVAC industry, and do we have to train our mechanical engineers in BIM?

6:00 PM  
Blogger Matt Dillon said...

Most definitely! With Revit MEP and the power of analysis based design tools combined with the information provided by the Building Information Model, MEP is one of the areas of greatest potential.

6:58 AM  
Blogger I am ME said...

I am a civil engineering student of Cochin University(CUSAT), Kerala, India. Thanks for the explanations, that gave a crystal clear idea about basics of BIM.
Traditional Construction Industry, which is in a fragmented form can be integrated by the use of such an Integrated model.

6:36 PM  
Blogger Unknown said...

Hi Matt,

This is still a very relevant topic for today in 2012 as it was in 2005. It really has not aged in it's topic and content.

4:23 PM  

Post a Comment

<< Home