Trimble’s John Bacus shares his thoughts on BIM’s dirty little secret, and why complex and expensive BIM tools are still basically being used for drafting
If you’re anything like me, for the better part of a decade you’ve been inundated with good news about how Building Information Modelling (BIM) is ushering in a new golden age of design and construction efficiency.
Certainly there have been significant projects built using the principles of BIM that could never have been built any other way. Frank Gehry’s Beekman Tower project in Manhattan is a particularly notable example. Through careful analysis and clever design optimisation, this beautiful building was also completed on time and on budget.
As my friend Ragnar Wessman (the ‘Father of Tekla Structures’) recently pointed out over a pint of Finnish lager, our industry still feels compelled to define BIM anew every time we talk about it… so we might as well begin there.
Back in the early 1990s, I found it pretty tough to get a job. The economy wasn’t great and as a young punk fresh out of architecture school, I was probably more of a liability than an asset to any firm that might hire me. The one thing I had going for me was that I knew about ‘computers’, particularly those made by Apple and, most especially, about doing computer graphics with them. And so, through what in hindsight seems an unusual turn of events, I found myself with a job offer from a prestigious design firm in Berlin.
Among the many things I love about the Germanic mindset is the belief that if there is a correct way to do something, then there really isn’t any reason why you would want to do it any other way. While 2D CAD systems were pretty well integrated into most architectural practices at the time, the tool of choice in Berlin was something better. Rather than doing the mental gymnastics necessary to project 3D building designs into 2D orthographic drawings, it was possible to make a fully detailed 3D model of the project and then let the computer make your drawings for you. To that end, we used ArchiCAD, running on the hottest Macintosh computers we could find.
ArchiCAD was great, something different from the rest of the 2D CAD systems of the day. We could build complete ‘Virtual Buildings’ with it: 3D models with building objects in them that were more than just simple line drawings that architects had been creating by hand for a thousand years or more.
Virtual building, which seems indistinguishable from what we think of as ‘BIM’ today, promised that you could work through the entire design of your building to completely constructible levels of detail in the computer. All the troubling and difficult problems that the design team might encounter in the field could be found and resolved virtually before the first shovel broke ground on the jobsite. It is now common knowledge that decisions made during the design phase of a building are much more cost effective than decisions made after the concrete has started curing. Additionally, decisions made during design are typically better, more thorough and carefully considered. Design improves with simulation, analysis and iteration. BIM permits that when effectively implemented. Nobody likes to discover design problems on site when there is no time to consider alternate solutions.
But there’s a bit of a dirty secret hiding in the rhetoric. Time and again as I’ve talked with friends and colleagues across the AEC industry, I find the same basic patterns. The complex and expensive BIM tools they have licensed for their teams to use are still basically being used for drafting. Flip those great BIM projects into a 3D view and you’ll find a hodgepodge of geometry that was really just put in place to make the drafting look good. In the spirit of full transparency, I was guilty of doing exactly the same thing all those years ago in Berlin.
If you think of drawings as being the deliverable that gets the building built, then this all makes terrific sense. Why waste billable time building a detailed model that doesn’t contribute to getting your story across any better than a solid set of drawings? Which, by the way, is what an architect’s contract says they are obligated to deliver. Anything more just seems like a lot of wasted effort.
If all you really want from BIM is a way to automate drafting, then there are dozens of tools on the market that can convert polygons in a model into lines in a drawing. Even SketchUp, which we designed to be the simplest tool we could imagine would be useful to an architect, can pull that off without too much fuss. But this is where we have to start thinking differently. There really is a better way, and it doesn’t have to be all that complicated, either.
The real benefits to BIM come from simulating a project to sufficient levels of detail as to remove uncertainty from the construction process. We should be able to know exactly how the project is going to look, how it will go together and how it will perform. We should also know how much it will cost and how long it will take to build. And experience now shows that we can simulate all of those things to a high degree of certainty in the computer before construction begins.
That simulation can and should happen throughout the design process, starting right from the first conceptual sketches. There’s no reason to wait until the design has grown into a highly detailed state before testing it for constructability, affordability and performance. In fact, the best and most impactful design decisions are almost always made during the conceptual phase. That’s when the design is still plastic enough to support big moves. Later in the process, everything is so locked down that you’re often doing little more than rationalising your earlier decisions.
If you accept that pre-construction simulation is where the real benefits of BIM exist, then there’s a really simple story to be told. Because then, at its heart, BIM is a methodology; not a tool. 3D modelling (detailed or conceptual), energy analysis, drafting automation and parametric object definitions all have their place in a BIM process. But BIM is defined essentially by none of them. You can do very effective constructible simulations of a building using many different applications. Some people do so using only a spreadsheet.
If there is one defining characteristic of tools that support BIM processes, I think it is their ability to carry non-graphical attribute data on objects in the model. In other words, if you make a model that displays a set of columns, those columns should both look like columns (graphically) and know that they are semantically of a ‘column’ type. Typically, this means that they carry classifications and attributes, like type (Corinthian), height, weight, manufacturer, material, fire rating, etc., arrayed in some pre-arranged schema like IFC.
Models with attributes can be used for analysis (cost, schedule, energy use, code compliance, etc.) that support detailed design decision making, ultimately allowing the building to reach a fully constructible level of detail before construction begins. This analysis can begin at the first moment of semantic awareness. A simple polygon drawn on the ground plane can be tagged as ‘is:parking’ and immediately enter into conceptual cost analysis. As that polygon is refined over time, the conceptual cost estimate can also improve. But having that information early and right at your fingertips is absolutely invaluable.
So maybe there’s a case for a really simple BIM that, in the end, provides the greatest possible benefit to our industry. Forget all the fluff around this central point: BIM contains building objects that have properties and attributes beyond their visual appearance. That’s it. They don’t have to be fixed in a particular schema, they don’t have to have parametric properties, and they don’t have to be anything beyond the best possible representation of the design at any given moment in its development. Loose when the design is conceptual; increasingly refined as the design iterates over time.
Not coincidently this is how Gehry Partners uses Digital Project, the AEC industry-specific implementation of Catia that Gehry Technologies built and has used on so many truly groundbreaking projects. It is also pretty much the same system that we call ‘information modelling’ in SketchUp, allowing any schema for object classification and any dictionary of attributes to define objects in the most appropriate semantic for any given project.
The trouble with many of the big BIM tools that apparently dominate the popular BIM discourse today is that they try to do too much and, in doing so, they overly and prematurely constrain the design process. Their advanced parametric tool boxes are full of wall tools, slab tools, window tools, spiral ramp tools, etc. They excel at making building information models, as long as your building is made of parts they have already imagined you would need to draw.
Tools that don’t impose such rigid schema on the design are much more conducive to making great design decisions. You should be able to impose your own schema on the design to serve your project’s particular requirements. Want to reimagine what it means to enclose space in your building using something between a wall and a slab? With a properly unconstrained system, all you have to do is draw it the way you imagine building it.
When I asked Ragnar to define the future of BIM over beers that night in Helsinki, he said that the future of BIM was to disappear; meaning that it would eventually be replaced simply by, “the way we make buildings”. I think he’s exactly right. In our zeal as technologists, my CAD industry peers and I are always looking to define something groundbreaking, revolutionary and new for the markets we serve. Maybe this time what we really need to acknowledge is that the revolution, with all its high-minded rhetoric, is largely behind us.
John Bacus is product management director of SketchUp, Trimble Buildings, where he is responsible primarily for the growing SketchUp family of products.
If you enjoyed this article, subscribe to AEC Magazine for FREE