Friday, March 1, 2013
In 1995 Gartner Inc. gave us this clever graph which illustrates the way technologies are perceived in popular culture. The wikipedia entry for Gartner's Hype Cycle has a second version of this graph with various technologies plotted along its slope if you want to get an idea about how this works in general. In relation to BIM, we are sliding into the valley of disillusionment. This has little to do with the technology itself though. It has everything to do with how it has been marketed. Whatever Autodesk's marketing strategy, they sold BIM like pop-music to architects and promised real advantages to using the process. They sold it with shiny graphics and data-views so dense with information that everyone got a buzz and rushed off to order seats of Revit in droves without understanding how their workflows would be affected to get to that point. Of course it takes hard work to make such a huge change in the way we work!! Doesn't anyone still remember all those late studio nights in college? You didn't get yourself to the place you are now without hard work, why would you be able to go further without it?
So, the design industry has climbed to the Peak of Inflated Expectations and begun to bitterly slide down the other side. The interesting part of this trend is that we've been told that firms that fail to adopt BIM will become obsolete and irrelevant in short order... and they will. The quirk is that we all believed that it would happen on the way to the peak while everyone frantically trained their minions to use Revit so they could say they were using BIM. But you can't make your minions do BIM without being involved yourself. That's the old assembly-line design firm mentality - "I'll sit in my office and make cool pictures while my project architect figures it out and his drafters draw the construction documents... then as Phil Bernstein likes to say, we'll dare the contractors to build it." We began our descent when design firm leadership realized that either one, this BIM thing wasn't delivering, or two, that it wasn't going to happen that easily. I have a feeling that many firms are on the Slope of Enlightenment now -- accepting the challenge I posited in my tepid rant from yesterday.
I alluded to the exciting BIM uses I've seen from contractors and owners over the last year or two, and I think they are getting more out of BIM now because the flashy marketing blitz didn't hit them in their weak spot the way it did for designers. They responded better to practical matters like scheduling, logistics, management, budgets. They expect to get down in the trenches and muck out some solid processes that have a return on their investment. Personally, I'm a paradox. The best professors I've ever had have explained architecture as inherently contradictory - simultaneously asking us to think like kindergarteners and grad students, accountants and painters, scientists and mad-men. I like that role. I took to computers early in 1984 and started frustrating english teachers with word-processed papers that cut corners off the tedious procession of note-cards, outlines, rough drafts... But in the 1990s, the University of Kentucky had almost no serious ambitions about CAD or computers and we hand-drafted and built in chipboard with Exacto knives and Elmers glue. I learned architectural computing after my degree which may have helped me accept that improvements aren't a given. In other words, my expectations were not, and ARE not inflated. I slog it out every day looking for chances to make that model more useful. Make it go further. If contractors complain about the way architects make models, I try to improve that. If owners want to explore the ratios of rentable square-footage to common areas in relation to their pro forma, then I want my model to react to that... and not just once. I want that to be an option every time. So if I need to change the way I build things, then that's OK. Another great truism I've added to my arsenal is Phil Read's BEER test for BIM innovations: Better, Efficient, Elegant, Repeatable - the hook being that any innovation must be measured by how much time it saves us so we can drink beer on Friday instead of working overtime.
Heading up that slope of enlightenment to the plateau of productivity is going to take thinking like that. The problem for design firms is that the enmeshed ideas about hierarchy are currently inhibiting that thinking. The drafting minions tasked with doing BIM are under-educated in the high-level thinking that makes you an executive level designer and the leadership lacks (and looks down on) the computational literacy that their drafting minions graduate with. If you work for a firm where the partners have not had at least introductory training in the software you are asked to use, then you aren't going to make it far "doing BIM." On an end note, I commented to Phil Bernstein at the IPD and BIM Symposium in Charlotte last month that we seem to be saying the same things we said to architects back in 2005. To which he responded, "the same things still need to be said."
Thursday, February 28, 2013
This is an interesting question and the answer is not what most architects and engineers expect. In most of the sales pitches for Building Information Modeling we hear about the benefits to designers that come with increased accuracy, greater efficiency, more computing power to design more energy efficient buildings… and of course the 3D pictures are very sexy. But BIM in the trenches is greatly underused (and greatly misunderstood) in the design half of the construction industry. While the McGraw-Hill reports that BIM adoption among architects is at 71% in 2012, this is based largely on the architect’s perception of “adoption”. In my experience, any architect that has purchased a seat of Revit will claim to be engaging in Building Information Modeling. That makes for impressive percentages but does not reflect the truth of the situation. What are architects doing with BIM? That’s an interesting question… we’ll get to that in a second.
On the other side of the design wall, the adoption rate for contractors is even higher (74%) and they seem to be getting more out of their software. The primary uses of BIM in construction are job site planning and logistics, and analyzing the constructability of designs. I think this is a fascinating development because while architects have been running from liability, contractors are learning to capitalize on a fragmented design environment that doesn’t take a stand, and with BIM they are doing it more efficiently because they know much more quickly and with greater accuracy exactly where the shortcomings in design team documentation are. This of course leads to change orders later on and the contractor can plan on those now. BIM gives contractors more control and helps them manage the unexpected faster and earlier in their work which saves them money. If I’ve learned one thing in the past year of working with contractors on their side of the industry, it’s that the design side is a mess.
Even when architects use BIM authoring software, they don’t seem to be able to make anything useful with it. Their consultants (if they do use the same software) still produce models and/or drawings which conflict with the architect’s design. I actually spent a whole week writing an RFI for just inconsistencies in types, numbers, and orientations of light fixtures between the electrical engineer’s plans and the architect’s Revit model. That’s a long time, but it was another whole week before a response came to clarify the layouts. During those two weeks, nothing was ordered and nothing was installed, nothing could even be modeled for coordination with other trades. What a waste! Two Weeks! So, if the design teams aren’t coordinating their designs with BIM what are they doing?
Well, from what I can tell, the majority are making nice renderings, and complex families to lighten the drafting load and demonstrate formal gymnastics. By default, they have to use it to better coordinate their internal construction documents because it’s built into most software packages. Beyond that, the very savvy are doing schematic daylighting studies, some energy analysis. The very best users I’ve come across are leveraging their BIM data to do LEED analysis and structural design. It’s very boring. I completely glaze over when presented with yet another fancy parametric skin. It’s ubiquitous now… Students can do as much with photo paper and binder clips. The making of forms in design has met with the same dilemma painting met once the camera was perfected. I’ll be honest, the most exciting uses of the technology are coming from the contractors: prefabrication, laser scanning, RFID chips, augmented reality, robotic surveying, cloud based project communication… Where design teams are lucky enough to be involved in a non-traditional procurement environment, they have access to much more innovation than when nestled snuggly in their design silos and honestly, I don’t see them coming out anytime soon unless forced to do so by owners. For designers, everything is too expensive, too risky, impossible… of course it isn’t!
Here’s what they are missing out on: using their BIM data to check code compliances, high-level space planning, rapid design testing for energy efficiency and day-lighting, very early estimates with quantities that can be monitored during all phases of the design process, and much more. I’m always very excited to discover new jobs for the data we put into our models. Building in schedules, filters, plans, keys, and families that not only check for code issues, but also help document compliance for plan reviewers is just one suite of tools I use on every project without add-on software. It comes down to understanding what you have in front of you and knowing how to leverage it. These are not graphic tasks and while architects in particular disapprove of any task that doesn’t put ink on paper, they have a huge return. Spending one day at the very beginning of your project setting up the environment you will model in (location, elevation, orientation…) you will save yourself weeks of headache while coordinating with consultants and cut hours off hundreds of tasks that will inevitably arise during design. I spent an day recently tweaking my material library (which is now transferable in Revit 2013) to improve estimating. All I did was make materials to differentiate between specific sizes and spacings of studs in my walls (e.g. 3 5/8” Metal Studs (16”O.C.) so that I could schedule quantities of different walls rather than having everything show up as “Metal Stud Layer” regardless of its construction. Not a big deal but it does give me estimating power. Power which is now available at the first phases of design if I want it because it’s built into objects I re-use. Having a filter that turns rated objects a certain color in a plan view seems simple but if you’re making sure that doors in walls have the right rating… well, it saves a lot of time.
So rather than let this be a rant, let’s make it a challenge. I’d love to see elegant BIM uses that are not predominantly concerned with the generation of forms. We can all make cool looking stuff. Very few of us will ever get to make much cool stuff professionally. But we all have to be good managers of our jobs and designs. Think about what you hate doing, what goes wrong or gets missed with annoying frequency and think about what little changes you can make to your workflow to eliminate those annoyances, save you time, and save money!
Friday, January 4, 2013
At the beginning of the summer, I wrote a post about type catalogs for families with a lot of variations on a similar design. I liked the type catalog because it let me build a very large database of types quickly without having to make a new type each time in the family editor. But I am considering another benefit to the type catalog approach.
A friend of mine has asked for some coaching on the making of some casework families. As I was considering her request, I found myself thinking back to those laboratory casework families and planning an approach to avoid that mess. However, I was also thinking about the hundreds of unused families I purged out of an architectural design model the day before. I suddenly realized that I did not have the same issue with the structural model I was given. Why was that? Well, most structural families are loaded via a type catalog so you simply select the members you need and load those. The result: a lighter model file. I know type catalogs are a bit daunting to tackle but, there are huge benefits to be gained by using them. The first of course is control of file size. You won't have to load in one hundred types with the family if you are going to use a handful in one area. Second, making new types and modifying parameters is going to be easier as time goes on because you can do it with Excel rather than modifying each type and applying the changes in the family editor. As long as the parameters don't change, you won't even need to use the family editor to make changes to a family. So, from a management point of view the family library is a bit easier to manage.
Look back at the Type Catalogs! post for the process using Excel and CSV files. Also, I learned how to do it from Michelle Van Kolck’s blog on Type Catalogs at BIMBoomBam.
GRIPE OF THE DAY: If you are using Worksets, make sure you are working on the right one! It will come back to haunt you later.
Saturday, December 15, 2012
Americans have a curious fetish worship when it comes to tools. We seem to think that possession of a tool automatically gives us some status. Everyone knows the type. Buying photoshop doesn't make you a graphic designer, a super powerful computer doesn't make you a elite gamer, a box of records doesn't make you a DJ, and buying a car doesn't make you a awesome driver. But everything is sold to us that way... and we buy it. It's rampant in our culture and perhaps it is a global phenomenon. The fact is, while a tool provides the possibility for some evolution of characteristics, it does not automatically bestow them. When this infringes on our own expertise we understand it plainly. The same people who scoff at some new designer saying things like, “…they think they can buy a box of prismacolors and suddenly they’re a designer!” will turn right around and say, “We bought Revit so we’re doing BIM!” Without effort, no one becomes anything by owning a tool. A tool simply provides a means to express that which we are becoming. The use of the tool manifested by the owner is wholly a product of the owner and the cultural framework he or she exists in. So a culture that has thrived on siloing information, giving orders without participating in their execution, and covering their ass, isn’t going to easily embrace the collaborative, open design effort that is characteristic of a BIM project. So they aren’t going to be doing Building Information Modeling, they’ll be using Revit to make drawings.
The most frustrating moment I’ve had in all my years of using Revit was an encounter with a design principle who liked to hole up with his drafting board and draw sketches by hand. It was easy enough to model whatever he had sketched and Revit was immediately providing design feedback as it should. However, the design principle became agitated by so many design concerns in the early phases of the project. “It’s too early to be making models!” he’d say. “We need to design first!” he’d say. “We can Revitize it later!” Well, this is what I mean about cultural limitations to using a tool. We have a designer that doesn’t understand why we’re making models. We have a designer that doesn’t realize that the model is a design tool. We have a designer that thinks of models in one narrow way. In this case the intent of a model was a presentation graphic and a stack of paper drawings. Can you do Building Information Modeling in that environment? Not really. Regardless of your deliverable, BIM provides a team with information about the building. It’s a simulation. The same way aircraft designers build virtual models to simulate performance and build-ability even though their deliverable is a plane, we make models to give us more information to design with.
We, very deliberately, insist on asking questions before we’ve finished with design. Why wait to ask about a problem you can see immediately just because it isn’t the time we usually know about it. Why waste effort on a design solution that you already know has issues? The longer you wait the more it will cost you to face the issue. I almost hate to show the McLeamy curve developed by HOK (above) because it should be a part of architectural culture by now - blasé. Doing Building Information Modeling inclines you ask design questions early in the process where decisions and changes can be made at low cost with minimal collateral change. It’s never too early to begin modeling. By the time you start making contract documents, you should already have a refined idea of how the building will be put together. You don’t grasp that from a drafting board in an office away from the model and the team building it.
I’ve been very interested in tools and human culture since architecture school in the 90s. Tools tend to influence how we view the world and how we view the world tends to influence our tools. Revit is only one of the tools being dropped on Architecture, Engineering and Construction. How will they change our cultural context and processes as we move forward? Some of this comes from a casual post I made on Facebook this morning not realizing that if I hadn’t been taking care of sick children I’d have been bombarded with news about the awful events in Connecticut. The discussion of tools apparently has a very wide reach. Why shouldn’t it though? Tools essentially make us human. I cannot think of a single human that does anything without a tool… even if that tool is hand or a voice. In each case, we suddenly find ourselves confronted with some object and we begin finding uses for it. The first hominids walking upright must have been puzzled by suddenly having these two grabbing things dangling beside them unused. What to do with them? Some hominids were perhaps more imaginative than others in the uses they put them to and those moved forward into human history. We suddenly find ourselves with these computer applications at our disposal and some of us will use them to do the same things we’ve always done - perhaps a little quicker than we used to. Others will find very imaginative uses for them and move forward.
Friday, December 7, 2012
The first thing I like to do when I get a new machine is to get the biggest model I have, turn everything on, open a 3D view and move around. I have yet to be presented with a machine that handles this task gracefully. You can spend $5,000 on a machine built for Revit and you won’t get that fantasy to come true. However, there are some decent strategies you can use when modeling that will help even a modest machine handle Revit with more grace. Some strategies are just good sense measures for planning and others are big No-Nos that bog down Revit and increase the amount of computing that needs to be done. So here are a few gems:
- Be judicious with Linked Files.
This seems to be the first thing people try when file size overwhelms them. In my experience, taking a model and breaking it up into multiple models for the same discipline is an invitation to disaster. I recommend using worksets to accomplish this same task. This, of course, means that some planning needs to be done at the beginning of the project to decide how the model will be broken up into worksets. I agree with strategy of dividing a model according to the way it will be worked on. I usually, divide out the Exterior Shell, Vertical Circulation, and groups of floors (depending on the size and complexity of the project.) You may not realize that you can select which worksets will load up when you open a file. Click the little arrow to the right of the Open button in the Open dialog box and select “Specify…” This brings up the Opening Worksets dialog box and lets you select what parts of the model you want to load and work on. For example, if you need to work on the Exterior Shell, you simple select that workset and any others you might need to work on it (Shared Levels and Grids, Structural, etc..) You can skip loading the interior walls, elevator shafts, stairs, etc. and work on only what you need.
- Avoid Groups.
Groups, while they might seem to be efficient, are not at all. They differ dramatically from Component Families because Revit will still track the individual objects within them as well as data for each group and each instance of a group. This is because you can exclude elements form individual instances of groups. Rather than simplifying the model with them, you are actually dramatically increasing the amount of data Revit has to crunch. If you use Groups to layout repeating elements, leave them grouped only as long as necessary.
- Use Component Families instead of In-Place Families.
This goes along with the same reasoning as avoiding Groups. A component family provides the efficiency you want without weighing down Revit.
- Keep Templates (and Files) Lean.
Basically, the more stuff there is in a Revit file, the more your computer has to keep track of. Loading up your templates with hundreds of families that you might use will cost you. Keep your content lean. Delete component families you aren’t using. Get rid of excess views, reference planes, and other junk. My rule for Reference planes is that if you didn’t give it a name when you made it, it probably doesn’t need to be there.
- Hidden Line Views.
I haven’t tested this yet, but several people have told me that the Hidden Line view option is considerably slower to work in than a shaded or wireframe view.
- Put your Revit Links on Worksets.
This is a new one for me but I’ve tested it and it really does improve performance when combined with the selective loading of worksets. Basically when you are linking a Revit file into your Revit model, you should make a work set for it first and then activate that workset and load the link. Supposedly this allows for more complete unloading of links if you do not select their work set when opening a file. So, for example, if you were linking in a structural Revit file, you would go to the Worksets dialog box via the Collaborate tab and create a new workset for the Structural Link. Click OK. Then select that work set from the pull-down menu under Active Workset. Once that is completed, you can link the Revit file. Make sure you change the Active Workset back to the appropriate one after the link is complete.
- Limit DWGs in the Model.
I’ll be honest, I hate having DWGs in my models. Part of it stems from the fact that I resent wasting so much time in AutoCAD instead of working in 3D but in reality, DWGs slow down Revit a lot. All you have to do is open Visibility Graphics and go to the Imported Categories tab to see how much garbage comes into Revit with a DWG. Even when I get CAD details from manufacturers, I tend to load them into a drafting view in a dummy model and Revitize them there, then bring them over to the real model as Revit data. You can junk up a model really fast with DWGs. Use only what you absolutely need. Also, NEVER… EVER… EVER… explode a DWG into grains of sand. At best a partial explode in a dummy model is reasonable - NEVER in your working model.
Hardware is a super hot topic among Revit users. Nothing gets them more talkative than a lively debate about their newest machines and what hardware makes a difference. You can look all over the internet for these discussions - and you should before shopping for a Revit machine.
The basics are easy enough though. The primary power source for Revit performance is processor power. Go to CPU Benchmark for the low-down on CPU performance. You’ll want something from the High-End CPU Chart for sure.
Get as much RAM as you can afford at the fastest speeds you can get. I’m not sure if the 20:1 ratio still applies in 2013, but its a good guideline. (20MB of RAM for each 1MB of file size.) Whatever Autodesk recommends for RAM is never enough. I am currently running with 32GB and I still wish I had more.
Be careful of the recommended systems on Autodesk’s website. I ran into an issue with the recommended graphics cards where the cards recommended were obsolete technologically, but cost more than cutting edge cards. While generally, the graphics card will not boost general performance, they do affect rendering speed and allow you to run multiple monitors at higher resolutions which maximizes the real estate available for Revit.
The hot new thing currently is the use of solid state drives. I have to attest that putting your OS and Applications on a SSD will blow your mind when it comes to load times. I keep my central files and libraries on a traditional drive but let Revit create local files on the SSD. This works out quite well and if SSDs do have the limited lifespan they are rumored to have then I can limit the amount of writing that gets done a little bit.
Lastly, if you don’t know someone who enjoys building high-end computers (a serious gamer, for example,) I suggest you make friends with some. A company like BOXX which makes excellent Revit machines will charge you as much as $6,000 for a high-end machine. If you find someone who enjoys the challenge and is willing to charge you for the materials and take a 12 pack of beer for the labor, you can spend about $2,000 for the same machine. Just make sure they understand what Revit wants from the beast you are asking them to build… A gamer tends to think that the graphics card is paramount rather than the CPU.
Management of ModelsJust a few quick notes about management of models: Be sure to review your warnings regularly. About once a week, click the Audit checkbox when opening a Central file to make sure everything is ship-shape. Change you link paths from relative to absolute and clean up linked files to get rid of unused family references and other junk. Something like 60% of file size is from family references. Keep your worksets organized and clean. I realize the monumental task this can be when you have larger teams working in a project, but it will be worth it to make sure the proper elements load (or don’t) when you select the worksets you want to open.
I’d love to get other suggestions. None of us can afford to keep getting new machines to keep up with the modeling loads and new releases of Revit each year. If your hardware is preventing you from getting the most out of your $4,000-$6,000 software then you aren’t getting the full benefit of using the software and you’re wasting money. If you can do a little bit of revit on your machine you may as well be using CAD. The more you embrace the BIM process, the more you’ll want to try energy modeling, day lighting, estimating, rendering… all the things that make BIM worthwhile. Don’t let your workflow and your hardware hold you back!