As I was writing my recent post on all the little changes in ARCHICAD 19, I was reminded of a conversation I had with Bradley Madison, the BIM Manager at Hennon Group Architects, PA. I don’t recall if this occurred via LinkedIn or at the 2015 GRAPHISOFT North America BIM Conference. I’m sure it was a little of both. We were discussing a video on Complex Profile Siding that I’d done (this one). He mentioned that he’d tried a similar technique, but that it didn’t work on the scale of project his firm does. We talked more. He mentioned that he had last tried this technique on ARCHICAD 16. I told him it was time to try Complex Profile siding again. So much had changed since ARCHICAD 16. My request is even more valid now that we have ARCHICAD 19 and all it’s speed improvements—at the time of our conversation, ARCHICAD 18 was the latest and greatest. This technique might still be too cumbersome or processing power intensive for the type of work his firm does. But neither Bradley nor I know for sure.

We all do our best to read about the latest version, watch all the new feature videos, and attend user groups right after the new version comes out. But that’s no guarantee we’ll notice all the changes that matter to our day to day work. The big marketing push won’t cover everything. GRAPHISOFT might not share a video on a particular feature for months. We might read about some new feature and then forget about it when we come across a situation that would be improved by the new method. Or we might have missed a development in ARCHICAD from a few versions ago. As a result, it’s always worth questioning workarounds and non-perfect workflows.

We do extensive tests to understand why files slow down or balloon in size. We catalog that knowledge and do what we can to follow best practices. We research and develop methods to improve our models. Sometimes these solutions work; sometimes they don’t. If they succeed we often include them into our processes and never revisit them. If they don’t work, we do something similar. We cast off these ideas as a good concept that just didn’t cut it—maybe our coworkers could never manage the complexity, maybe it just took too long, or maybe there’s something about ARCHICAD that made the genius solution almost perfect but falls short and therefor useless. Both those results—using a good method and discarding an unsuccessful process—should be periodically revisited. Good ideas decrease in value over time and the constant evolution of ARCHICAD means our knowledge and skill set is judged against a shifting set of best practices. Cever ideas become bad ideas. Bad ideas become good ideas. Great ideas that are impossible become good ideas that are possible, which is definitely moving in the right direction. Often these techniques become worth doing, not just because they become easy and lighter to use, but also because we can extract more value out of the results (more on that later).

The addition of IFC Mapping in ARCHICAD 18 caused one of these shifts for me last year. I started a project in ARCHICAD 17 and hit a wall of impossibility: I wanted to schedule bits of data uniformly across Columns, Beams, Shells, Roofs, and Slabs, but couldn’t. Once ARCHICAD 18 was released, I migrated the project and the impossible became reality. I still found stumbling blocks, but I had crossed a threshold. If you aren’t comfortable with IFC Mapping, here’s a mountain of information and videos for you.

Does that Still not Work?

Sometimes a lot of semi-invisible changes turn a bad solution to a good solution. In my example of Complex Profile siding, a convergence of forces have made that a more viable solution for ever larger projects. In ARCHICAD 17 Building Materials and Priority Junctions offered a solution that was lighter on processing power than Solid Element Operations (which was often needed with my technique), and they also made the siding even more intelligent, and therefore useful. In ARCHICAD 19, background processing and other speed enhancements means the same hardware can handle ever more complex models. I’m sure if I sat and thought I could find more improvements between ARCHICAD 16 and 19 that change the relationship between Complex Profile siding and model size. The details aren’t important, just the understanding that the complex and intertwined improvements of ARCHICAD result in workflows moving from crazy to crazy easy, sometimes gradually and other times abruptly. I would argue that in the case of Complex Profile siding, the changes are gradually aggregated as multiple improvements compound. Maybe the technique still isn’t ready for all project sizes. But let’s ask the question again next year, and the year after that. Dimension improvements are a good example of a feature evolving in discrete jumps. Back in 2012 (ARCHICAD 16), I wrote a blog post entitled Tips, Tricks, and Cheats… ArchiCAD Dimensions. Almost three years later, it is still one of my most popular videos and articles. I covered the nitty-gritty basics of using dimensions in ARCHICAD; all the accumulated wisdom I’d gathered since ARCHICAD 9. There are some pretty cool tricks in there and, if you haven’t watched the video, you definitely should. If you have new employees, it’s especially useful because it covers everything they need to know. Except that it doesn’t. Not anymore. The video was great for users up to ARCHICAD 16. And ARCHICAD 17. And ARCHICAD 18. But if you just watch my video and don’t dig deep into ARCHICAD 19, you will miss many great new additions. Such as automatic pointers to Dimension Text and multiple lines of text for dimensions. Or the older option to turn on a background fill for Dimension Text (which I think I missed in the video). None of these features are hard to find, but after years of using other solutions, we are often blind to the buttons and options right in front of us. These dimension improvements are about the question “does this still not work?” and the related question “is this still the best solution?”.

Not 2012s dimensions

Is this Still the Best Solution?

Tutorial videos are beautiful examples of the decay of expertise. We pause and share our knowledge, then those answers stay on the Internet statically attracting eyeballs. Unless we revisit and revise the videos, they can give a false recommendation of best practices. When I started my own ARCHICAD tutorial channel on YoutTube, my first two videos were on interior elevations. My techniques were clever and I was really proud of them. A few years later I learned more tricks (nothing that couldn’t have been done when I recorded the first videos) and did two updated videos. Here they are (if you have time, they are all worth watching):

  • Jared’s Interior Elevation Advice from 2011 (ARCHICAD 15): Parts 1 and 2
  • Jared’s Interior Elevation Advice from 2014 (ARCHICAD 17): Part 1 and Part 2

It’s interesting to look at these videos now, so many years later. If you listen to my 2011 advice, it’s all about visuals. I shared clever ways to create the graphics I wanted using fast 2D solutions. By 2014, I was using smart 3D solutions. My 2011 advice is good and clever and effective. And fast. But the 2014 advice is smarter and faster. And it also completes a line of thinking, allowing me to focus on other things. And that’s the goal. It’s not about getting perfect, pretty interior elevations—or more realistic 3D siding. The goal is to make things mundane and automatic so that we can focus on more important problems. Once upon a time people spent whole careers setting up interior elevations. Back in 2003 I worked for Gensler in Houston for nine months before my final year at Rice University. As an intern at Gensler, using a very well known 2D CAD program, I spent weeks drawing interior elevations. I would go so far as to say I spent the bulk of my nine months drawing interior elevations (though I did spend some time building a full scale mock up of a monument to a former president). Thousands of dollars and hundreds of man hours (I was young, so maybe boy hours is the more appropriate term) were spent drawing line after line. Even years later when I was using ARCHICAD 9 in 2006 and 2007, I was still doing that same process. By 2011 I was at least using a hybrid approach, which is what I shared in my first videos. The interior elevations were connected to the model, but there was still time spent crafting the graphics. I pressed on and searched for better solutions. By ARCHICAD 17 (and now in ARCHICAD 19), my time spent on interior elevations is approaching zero. There is some set up, but it’s global work that can be stored in a Favorite.When I was doing 2D elevations between 2003 and 2007, I wasn’t being dumb or lazy. I was doing the best with the tools at hand. I was finding ways to be more productive than anyone else around me. I was doing the things you’d expect—grids of interior elevations, clever use of alignments and construction lines, using groups and objects where ever possible, excessive use of copy and paste, reusable jigs, pulling lines off the floor plans…. I was a progressive CAD monkey, even when I was using ARCHICAD. The same goes for my advice in 2011. They were good techniques, ahead of what my coworkers were doing and better than other solutions I’d seen. The processes matched the goal: fast, beautiful, coordinated drawings. That was the benchmark. But the bar of expertise always rises and the cleverness of our solutions always degrades. Always.

Had I never questioned my best practices, I’d still be focusing on the wrong goal. In 2003 all the work was done in the view. By 2011 half the work was done in the View and the other half was done in the model. By 2014, 95-100% of the interior elevation graphics were created from Elements outside the View. But still at that point most or all of the annotation was done within the View. Now in 2015, automatic 3D is a given. The question is what can be done next? I want to continue shifting data away from living in the View. How much can be automatically labeled using the new, much more powerful Label tool? I’ll still need to go into the View to label what I want, but the dream is to get it to the point where I turn on the labels, pulling information from elsewhere in the model, rather than adding dumb text manually.

I want all my interior elevations to be fast, beautiful, coordinated, and documented with minimal direct input from me on the View level. I want to go to an interior elevation and manage data not create it.

I’ll end this post with the same image I started it with. It’s where my interior elevations go next. The graphics are 100% from the model. I could gripe about a few extra lines around the window, but I won’t. Because I don’t care. I see this interior elevation and think about how I put no time into the specific view. I’ve turned on a Label for three elements: a Door, a Wall, and a Beam (which is the baseboard). Each Label is set to automatically pull a piece of information from the element it’s related to: an ID, a Surface, or the name of the Complex Profile. The below image is just a mock up. My template will require some Attribute naming tweaks, but that’s easy. And actually since it’d be a global change, I could work on a project like this today and then update the Attributes when I’m ready. Right now if I turned on the Inside Surface of all my walls, instead of Gypsum Board, it’d say “Paint – White Titanium”, which is the name of my generic painted wall Surface. “Paint – White Titanium” could be shown a hundred or a thousand times throughout my project and with one update to the Surface name, it’d be “Gypsum Board” or “Primed and Painted, or Color – TBD” in an instant.

Interior Elevation 2015You might be thinking “Jared, when is your new dimension video and your next Interior Elevation video coming out? Show us your new processes!” I have no idea. Before 2016? I hope. I want to do some research and experimenting to make sure I don’t miss anything. Follow Graphisoft North America on Twitter to keep track of all the latest ArchiCAD advice and news.


  1. Eric Bobrow

    Jared –

    You always impress me with the depth of your thinking.

    As someone who has focused for 25 years on Best Practices for ArchiCAD (and named my comprehensive online training resource “The Best Practices Course”) I have a keen interest in this topic.

    You have just reminded me that I need to go through all of the curriculum and my training videos to bring them up to date.

    It’s sort of like the painting of the Golden Gate Bridge in San Francisco – whenever they finish painting it (with a rust preventive coating) they start in again from the other side. It never ends…

    Eric Bobrow

    • Jared Banks

      Thanks. Always happy to inspire.

  2. Rob

    the only issue I see with using “automatic labeling” coming off the materials is that you’d need to, from the onset, have established the correct spelling and syntax for the materials names. For example, since Archicad likes to organize names, objects, numbers etc…alphabetically and numerically, the names of the materials could never be mixed up. What do I mean by that? Well, if you are referring to “Paint-surfaces” then ALL the paint materials need to be named, or at least started with, the same format, such as “Paint-Stucco White” or “Paint-Brows”, etc….but once other users start to name the surfaces to their own liking, then the system fails.

  3. Rob

    it always seems to boil down to the following:
    “Policing” of BIM policies within an organization.
    It’s all fine and dandy if you work by yourself but how many of use “really” work in our own bubble these days with all the touting of “collaboration” these softwares are designed to do.

    I do like the approach of trying to “automate” as much as possible. That would be a definite (huge) time saver, especially as it relates to the “annotations” of the drawings.
    Love the articles and youtube channel by the way. Very insightful and most helpful.

    • Jared Banks

      Rob, these are great concerns. And it’s been so long since I’ve had to police more than one or two people that I don’t know how well it’d go. BUT I think it’d more about mindset. If ARCHICAD users can learn to understand the importance of Attributes, then it shouldn’t be a big deal. And ARCHICAD users should understand the importance of Attributes, so this should actually be self reinforcing. Whether or not people use automatic labels, all people in the firm should be using the same Building Materials for the same stuff. And the same fills and surfaces. Whether or not you’re linking to the data, if half the team is using the Framing Lumber BMat and the others are using Wood BMat for studs, then there’s a problem. If they all use the same thing, then it’s just about teaching them to use the tool. And turn on the label.

      And as for naming…to me it’s all placeholders. It’s okay if the drawings are labeled weirdly on day 1 or 20. What matters is consistency. If everything is linked properly, then Paint-Stucco White can be renamed to “Stucco, painted white w/ three coats” or whatever you want. And finding that information shouldn’t be hard because a user is seeing that written on a drawing, so it’s an easy thing to remember and find. So my viewpoint is actually to not worry about naming things correctly on day 1. Accept that they are named incorrectly on day 1. And that the names will evolve. At the start all the surfaces could be renamed (as you pick them) to be super generic (exterior finish, interior finish, metal, concrete, stone, etc.). Then just like our models slowly go from generic to specific, so can the naming of all the Attributes. I know it’s good to start with specific Attributes. And certain ones like BMats can be hyper specific from day 1. But things like Surfaces can be a lot more loose as they will continue to evolve throughout the project. And I feel it is more valuable to tag different elements with like Surfaces and update the Surface than continue to change which Surface is assigned to which element. ALTHOUGH, the new Surface Painter makes updating Surfaces amazingly easy. But that’s for another discussion.

      I’m not discounting the issues you raise because it will be a challenge to convince people to follow a new workflow. But I think the upside is too great. And if a coworker refuses to automate, well when it suddenly takes them much longer than everyone else to do their job, hopefully that is pointed out.

      bonus random thought: maybe the title BIM Manager should be updated to BIM Enforcer.


Submit a Comment

Your email address will not be published. Required fields are marked *