Design Collaboration: Some of the Parts are Greater than the Sum of the Holes
Jerry Laiserin

Way back in the glory days of the Internet revolution, say 1999, the integration of everything with everything else seemed inevitable. Since the NASDAQ collapse of April 2000 blew holes in the idea of inevitable integration, software survivors have shifted to peddling mere parts of solutions. Is this a concession of defeat, or a shrewd backstep toward conquering the market?

"A" for Effort
Market leader Autodesk has adopted a nine-fold strategy based on a three-by-three or Hollywood Square matrix populated with diverse tools to let designers, builders, and facility managers variously author, publish, and/or analyze their work. For example, the squares under the "author" column contain different tools for designers (Architectural Desktop-ADT), builders (Revit), and facility managers (third-party outfits Archibus and FIS). No hint of industry-wide Industry Foundation Classes (IFC's) or aecXML to unify these threads. Not even a mention of Autodesk's own DesignXML. The message seems to be: break a project down to the parts that you directly control, and choose the tools that best optimize your role within the building information lifecycle.

Lest anyone think this is a bad thing, understand that Autodesk-like the billion dollar leader in any market-got and maintains its position by carefully pacing the mainstream needs of mainstream customers. What those mainstream customers are telling Autodesk right now through their buying behavior is that the existing AEC culture, process, and legal environment do not allocate the benefits of large lifecycle improvements fairly to those who would incur the costs of generating those improvements. Few if any owners offer their engineers and architects fee incentives for delivering smarter project documentation and models that embody more useful information throughout the design-build-operate process.

This is one of the key reasons that Autodesk Revit is now a wholly-owned costing tool for builders, rather than the independently-owned killer-app for designers that it once promised to be. One wonders, though, if builders are going to create their own intelligent models based on geometry and data from designers' models, wouldn't Tririga's Intelligent Building System, with it's oh-so-clever "virtual areas," actually be the better way to go? Today's San Rafaelista equivalents to Cold War-era Kremlin watchers couldn't help but notice that a recent Autodesk-Tririga press release touted benefits to the design-build, hospitality, and retail markets-echoing Revit's marketing direction in the waning days of its independence-yet Autodesk's side of the Tririga alliance appears to focus only on ADT.

Attack of the Killer B's
Bentley Systems garners a tad more than one-fifth Autodesk's revenue, but often seems to generate five times the buzz. Partly, this is a consequence of size; partly a consequence of not (yet) being publicly held. Whatever the reason, Bentley's current mantra-"create, manage, publish"-stresses the ability to manipulate the same information in the same tools across the entire project lifecycle. Version 8 of Bentley's flagship Microstation is now positioned as a Windows-like platform, rather than a free-standing design tool. Portfolios or collections of applications on top of this platform serve the specific needs of various players throughout the process, yet enable everyone to operate on the same data within the same model.

However, digging a bit deeper into the sales pitch reveals an implication that, though the model and data are always there, it's still possible to just put one's head down and crank out documents, or estimates, or whatever the task may be at hand. It's not that Bentley has abandoned integration, so much as it has modularized its offerings in ways that serve the practical needs of mainstream customers without sacrificing the more avant features demanded by users closer to the early adopter edge of the technology lifecycle.

E Pluribus Unum
My cover story for the June 2000 ink-and-paper version of CADENCE magazine addresses "AEC Interoperability and the BLIS Project" (as with most monthly dead-tree media, that story was written months before and may not be on-line for months to come). The tale of BLIS and its relationship to the International Alliance for Interoperability or IAI is too long and convoluted to delve into here. Suffice it to say that BLIS relies on IAI IFCs version2.0 for data exchange among diverse design and analytic applications. The principal design toolmakers and tools involved in BLIS are Microsoft Visio and Graphisoft ArchiCAD.

Befitting ArchiCAD's larger market share outside North America, Graphisoft has plunged deeply into BLIS-ful interoperability. ArchiCAD's robust modeling and data-handling capabilities make it well-suited to this role, and the differences between European and Asian AEC industry cultures and those of North America tend to make project-spanning interoperability a more rewarding process for all participants. That said, the latest spin from BLIS demo-land is that "we don't have to move all the data and geometry-even though we can." In other words, now that BLIS has cultivated the garden of data unification and application integration, it's OK for individual project participants to tend their own little patches.

The Pitch and the Pendulum
Marketing pitches and technology memes oscillate with predictable frequency. Two years ago, integration was everything and standing alone was standing still. Today, there's "enough" integration that it's "OK" to stand alone. Two years from now we'll all be clambering aboard the integration train once again. Players in one CAD-adjacent niche may soon be crossing over to others (watch for outfits like Timberline to expand outside their "traditional" markets). Horizontal, enterprise vendors, such as OpenText are going vertical in markets like AEC. International design software makers such as Nemetschek have yet to be heard from on this swing of the pendulum.

The most ambitious folks in the interoperability camp are pushing for a true "virtual building" process, which we'll report on in a few weeks when we revisit Stanford's Center for Integrated Facility Design and Berkeley's Lawrence National Lab.

Finally, never underestimate the Finns. Beyond the bevy of Finnish participants in BLIS, a Finnish startup called Enterprixe, running on top of ADT or plain-vanilla AutoCAD, promises to do for AEC model-based, real-time, on-line collaboration what Infrasoft's Arenium already is doing to shake and stir up model-based, collaborative infrastructure design. In case you're thinking, "Why should I care about Scandinavian technology vendors?" just give me a ring some time on your market-leading Nokia or Ericsson mobile phone.

< back