Revit

Autodesk Inc.    (formerly developed by Revit Technology, Inc.)

Review in Cadence magazine AEC Tech News March 15, 2001 -- Revit 3

Leonid Raiz and Irwin Jungreis, both veterans of Parametric Technology Corporation, founded Revit Technology Corporation in 1997. Firms such as NBBJ of Seattle, McClier of Chicago, and Macomber Construction of Boston are involved in the final development of Revit.

comments by Prof. Geoffrey Moore Langdon, Architectural CADD Consultants

This is potentially a major move in the architectural CADD market that in any case will effect many CADD software developers in both positive and negative ways. Possibly even more importantly, this could be the final catalyst to crystalize the paradime shift to 3D model centric architectural CADD that the architectural profession has been teetering on the brink of for quite some time.

Even if the deal ultimately falls through (remember the Justice department saying no to Microsoft buying Inuit (QuickBooks)), this establishes a new price for Revit, some four times their previous valuation, which would enable them another fresh start on venture capitalization.

Assuming the deal does succeed, and yes, due to the current incompatibility of not only their respective company philosophies, but also software design approaches, their actual data structures and file formats, this deal reeks of being cooked up by Wall-Streeters not core programmers; there will be a long period of sifting out the concequences - just as there was with SoftDesk, LightScape, GenericCADD, and the other companies Autodesk has bought over the past.

In April of 2001 Autodesk showed that they were totally redesigning their ArchitecturalDesktop - adding multi-story ability, greater parametric intelligence, hugely simplified interface, and abandoning their "legacy" underpinnings stating definitively "no more Object Enablers for older versions of AutoCAD" and they showed off the tandem project of ArchitecturalStudio nee StudioDesk as an example of where they are headed. Carol Bartz at that time also said that the 3D parametric model based approach would be the new heart of Autodesk's efforts (enabling them to get into countless new markets, such as automatic structural engineering framing, hardware stores wood cut lists, and on and on).

From what they said, it seemed to me they were describing the path that ArchiCAD, Revit, and DataCAD Plus have taken, and that they would need to take ArchitecturalDesktop WAY beyond where it currently is, something which would take several years. Ultimately, though, the effect is that, say five years from now, the idea of DRAFTING something in 2D in AutoCAD, DataCAD, etc. would be obsolete, as our clients, engineers, interior designers, and builders would be requesting "usable" files in the form of smart 3D objects and intelligent parametric walls.

The old kaamaaina amung us may remember the failed attempt years ago of when CADKEY tried to create a core CADD engine, combining the best of DataCAD and the CADKEY mech softwares, with just "a different interface" for the architects and the mechanical engineers, and how that effort failed because on the most basic level they could not even agree on how to define a "line" (the engineers needed it to be capable of being bent to 3D freeform NURBS curves, and the architects needed it to recognize a "ground plane" and be capable of being a "wall" but still be simple). In a similar vein, the vastly different strategies going on at the core Autodesk ArchitecturalDesktop development, and Revit seem equally un-consolidatable. I believe though, that in this case, such consolidation is not necessary. Remember the ADT software designers are already looking for a new data structure to enable them to go the next step.

Also, the incompatibilities many people have mentioned, (and I too have mentioned in countless presentations and articles) between ADT objects and conventional AutoCAD DWG and Revit, is created artificially. The ARX definitions of ADT objects have been a proprietary secret mainly for competitive reasons - they don't want non-Autodesk companies to import them. Autodesk has not defended the reverse engineering of their DWG file format because it helps to establish them as a standard. That does not mean they have to give away their new stuff, and that is the way they have played it. Their AutoCAD 2002, for instance, can read in ADT objects just fine. Thus, once Revit is a wholely owned subsidiary, Revit too would probably gain the ability to read in (and possibly do even much more) ADT objects.

Thus the implications of that are huge. An easy to use software (think of facilities managers, contractors, spec writers, and many others who don't currently learn/use AutoCAD) with access to smart parametric 3D "AutoCAD" drawings, could indeed finally make the idea of Web drag-and-drop symbols (with attached specs) a commonplace reality.

This will all take a good deal of time to settle out (I would guess about 22 months).

Other software developers (Bentley, Graphisoft, DataCAD, Nemetschek, etc.) could do well if they just seize the day, and really push the benefits of their particular systems during this period of confusion in the AutoCAD / ArchitecturalDesktop / Revit world. This whole activity validates the model-centric approach of architectural CADD softwares.

The notion of spending hundreds of millions to kill off a company with, currently, insignificant market share, is absurd. Autodesk obviously is after the technology, the only question is how the shift will be marketed. I imagine that they will try to make the shift reasonably "transparent" and that, assuming they keep the name ArchitecturalDesktop, that the engine behind the scenes when you open ArchitecturalDesktop circa 2004 will actually be what used to be Revit, with also some bells and whistles from the previous ADT efforts. On the other side, Revit users also will see the same Revit interface for the next few years - renamed Autodesk Revit of course - and will benefit from greater AutoCAD file translators. Ultimately though, it will just blend into one Autodesk product offered to both former ADT users and former Revit users.

The extreme rhetoric Revit has had in their previous marketing "be different, don't use old technology" and so on, casting AutoCAD as the bad way of doing things, which attracted a number of their passionate users, will still be valid - as the old 2D AutoCAD drafting methods will be still be what AutoRevit does NOT do. The web pages so critical of both Autodesk and AutoCAD have suddenly dissapeared though from the Revit site. Anyway, just as Autodesk ArchitecturalDesktop users have been trying to convince their fellow AutoCAD colleagues there is a better way, that effort, boosted with Revit, will continue.

Added notes 4/4/02 :
The full aquisition has been finalized so supposition about added venture capitalist funds in case the deal fell through are now moot.  Autodesk yesterday issued an extensive Q&A about the future of Revit, AutoCAD, ArchitecturalDesktop, and ArchitecturalStudio (at http://usa.autodesk.com/adsk/files/1911309_Revit_Close_FAQ_final2.pdf)

This leaves Autodesk with quite an an assortment of offerings :
Autodesk® Architectural Studio for freehand conceptual design creation and real-time multimedia design collaboration
AutoCAD® and AutoCAD LT® for production drafting of 2D construction documents
Autodesk® Architectural Desktop, Autodesk® Building Mechanical, and Autodesk® Building Electrical for increased productivity in detailed design development and construction documentation through an intelligent object-oriented modeling approach within a familiar AutoCAD environment.
Revit™ for process improvement through an integrated parametric building model
Autodesk® VIZ for advanced visualization of design concepts at any stage in the process.
AutodeskBuzzsaw services for effectively managing all building project information online in one secure location.

Autodesk says that they will continue ALL of the various products they have in the works, despite some considerable overlap, with the comment that different types of firms will choose the appropriate for "their type of practice and work process".  Currently, there are lots of incompatibilities between all these products - despite the fact that you will never see the terms "incompatible" on the Autodesk web site.  Well known, for instance, is the fact that once you use any architectural elements in the ArchitecturalDesktop flavor of AutoCAD (wall, window, door, stair, etc.) those elements will not be visible in several other flavors of AutoCAD and may even make the DWG file entirely unusable to users of, say AutoCAD 14.  The same is true of Revit objects (i.e. a "wall with a door in it" does not come through to ArchitecturalDesktop AS a "wall with a door in it" or vice versa,  only as lines), similarly there are issues with ArchitecturalStudio, ACIS 3D solids (polyline extrudes), and much more.
Essentially all this stuff is in an entire re-design mode right now.  Only AutoCAD 2D is fully mature and stable, but Autodesk has announced that they wish to move away from the 2D paradime and into "3D model centric building" which allows for artificial intelligence and "smart" add-ons such as Building Mechanical and  Building Electrical (which, of course, ONLY work with ADT models), thus it is as though Autodesk is starting a whole new "undiscovered country" with a mish-mosh of various programs.

We envision that this is going to take several years for them to sort out, let alone try to meld into any one (or more) coherent product.  The acquisition of Softdesk (developers of AutoArchitect, etc.) which, like this buy out, also took even the people within the company by surprise, took some 10 months for some people to figure out if they still had a job and who they would answer to (remember that AutoArchitect was a casualty of that buy-out, with only a few bits, temporarily, being incorporated into ADT 1).  The Revit acquistion is vastly more problematic than the Softdesk one (totally different file structures, not just a third party add-on to AutoCAD) and will take longer.  Reconciling Revit and ADT with what the new ArchitecturalStudio software is doing only complicates the issue further with whole software design teams having different cultures, philosophies, strategies, and future plans.
Our guess is that they will develop a series of translators for the array of softwares, with lots of sidebar notes about what info may be lost when you hit SaveAs to a particular sister product.  We also believe that they will be gleaning certain technologies from one of these to another to solve problems - Revit's file structure solves the problem ADT developers had in allowing for full multi-story buildings, Revit't parametric 3D modeler may help in the development move away from ACIS (for Inventor, AutoCAD, and Viz), ArchitecturalStudio's re-think of interface may eventually simplify portions of ArchitecturalDesktop and even AutoCAD.

Thus, years and years of development and changes, lots of profitable classes for AutoCAD training classes to teach and lots of upgrades.  Autodesk will benefit from the fact that large architectural engineering firms are extremely slow to change (the vast majority still use AutoCAD 14, an 8 year old version now completely unsupported by Autodesk) and will introduce these changes as improvements over time.  Revit-ers and ArchitecturalDesktopers will thus remain as small esoteric and exotic "AutoCAD-ish" minority, with their respective, seperate, user groups touting the benefits of their particular Autodesk flavor.
Our guess is that, right now, Autodesk has no clue which of these products will pull out as the ultimate replacement engine for AutoCAD (or even whether it is possible to change AutoCAD itself into a 3D parametric-friendly software) so they are supporting them all.