Friday, February 15, 2013

Time for a Revit Revolution

Have you heard the news?! Autodesk has a new app to help you model monsters! (Cost: $2 introductory pricing) It seems pretty obvious these days that big A's attention has shifted away from core BIM tools and towards mobile app development.
They are very proud to tout over 12 million professional customers, and in a recent Forbes article, “Autodesk’s Brilliant Customer Strategy” - they claim a new concept for development:
“Autodesk releases its new products first to consumers, thus turning the product vetting process upside down. Consumers pay less but they expect more. They can be far from polite or patient, and will only tolerate very short learning times—and few bugs—in new, untried products.”
Sure, Autodesk has developed some interesting mobile apps for our industry such as BIM 360 and FormIt, but lets take a look at their main BIM platform – Revit (Cost: ~$6,000 – $13,000).
When Revit was introduced to the market around 2000, it was ground breaking. Never before did the architecture industry have a database-driven, parametric modeling tool that was relatively easy to use. In the first few years of existence, Revit Technology Corp made vast improvements directly related to customer feedback sessions they would host annually, right after they'd have a major version release.
Then Autodesk took notice and acquired Revit in April 2002 for $133 million dollars. Since then, it seems there really haven't been the kind of innovative improvements we saw in the pre-acquisition years. Granted, those years were really about taking a brand new platform and getting it up to speed for the larger market of AEC users, but it seemed like the trajectory for innovation was much more steep than it has been in the years since.
The Revit development team often uses the term "fit and finish" referring to commands, tools, and features they initiated, but really haven't improved to work the way they really should. Take the fairly recent addition of the Parts functionality. This feature was intended to allow builders to take a design model at consists of singular assemblies such as walls, floors, and roofs and break them down into individual components (finish, substrate, structure, ...) for more accurate phase scheduling, estimating, and so on. Great! So, the builders can link in the architectural model and...uh, no. It doesn't work on linked models. Really? [CORRECTION: You can create parts from linked models in Revit 2013, just use the Tab key to select a model element within the linked file and the Create Parts button will become active.]
Allow me to review some specific issues my industry colleagues and I have been asking for over the past few years – a few of the larger ‘fit and finish’ issues that have yet to be resolved in my opinion:
  • Faster UI - While I believe Revit still has the lowest learning curve of all the BIM software on the market today, it's user interface still seems quite slow. Ever since the ‘drunken leprechaun’ debacle with the 2010 release of the new ribbon interface, it just seems that I'm always waiting multiple seconds for the ribbon and command buttons to react - even with a blank project file open. As a customer paying $6,000 and up, I would expect the basic software to function crisply and cleanly. As the Forbes article quotes about the app users, “Consumers pay less, but they expect more.” So, should the inverse be true…that we (I guess we’re not really ‘consumers’…) pay more, therefore, we should expect less?
  • Handle larger projects - This is another area in which Autodesk says that performance is always their top priority. They have made moderate strides in improving performance of large projects, but no real game-changers. They have improved the functionality and workflow for linking models, but certain quirks still remain such as walls not joining between linked projects, and it's still a pain in the butt to manage view references between linked projects.
  • Schedule templates - Here's one I was hoping would happen sooner because of the functionality Revit has for MEP related to panel schedules. The scenario is for buildings in which you need to create a series of schedules that are usually based on each level. For example, if you need to generate an area schedule for each level to include on code analysis sheets. To do this in Revit, you need to create a unique schedule for EVERY level - essentially using a filter to show only Level 1, Level 2, Level 3, and so on. Why can't we create one template schedule (kind of like a Keynote Legend...hint, hint) that is then placed on a sheet and will report only the areas visible on that sheet. Could be a huge time saver for firms working on mid-rise and high-rise buildings...but alas we must trudge ahead with the same functionality we've had since 2004.
  • Better custom/complex modeling - We've seen some interesting developments get infused into Revit related to their conceptual design environment, but there are still some annoying gaps that prevent truly complex forms from being developed through to construction documentation and fabrication. Why can't solids automatically join and heal? Why does a complex curtain wall system need to be created as a mass and then be inserted into the project? I'm simplifying a complex discussion, but I think this workflow could be vastly improved.
These are just a few areas where I feel Revit could be improved, but they have been at the top of my wish list for years. Perhaps it is time for Autodesk to let Revit go to a company who cares about true innovation for the users who are actually designing, constructing and managing buildings. I'm pretty confident we'll never see "Bentley Revit" or "Graphisoft Revit" - but at one point I was inspired by the possibility of "Google Revit." Given the latest trends, maybe the next idea will sound something like "Trimble Revit." All I'm saying is that it's about time a once innovative platform is returned to the status of hands-down game-changer.