The ARCH

Are browser-based virtual environments the way of the future for virtual design visualization?   Here is a brief overview of 4 of the many emerging platforms or related services to watch, and some thoughts about how they fit into the bigger picture of virtual worlds and architecture.

I’ll start with MellaniuM.  I’ve been following MellaniuM’s work with the Unreal platform for quite a while now, and continue to be impressed,  especially with some of their newest demos using Nortel’s web.alive platform.    In addition to architectural visualization, they offer many additional application areas including GIS, education, medicine, entertainment, archeology, film making and more.

MellaniuM can take CAD models, and import them into the Unreal game engine where they tweak and refine the build to satisfaction.  Unreal includes a built-in editor (I found a video of that editor here), but it isn’t anything like SL’s building tools and probably not very useful as a modeling environment.  After they’ve situated the build properly inside of Unreal, they turn it over to Nortel – who then enables browser-based realtime access to the model.  Here is a case study of an apartment building imported by MellaniuM:

I was very impressed with my first experience inside the web.alive platform after visiting one of their first builds.  It was eLounge for Lenovo (accessed here).  Before my visit, I had to install a plug-in, and download the environment (totaling over 80 mb – not exactly ‘thin’), but found the installation process to be relatively seamless.  It did freeze my browser twice before I got it to work, and I had to close my other apps so it would run faster, but it wasn’t a show stopper by any means.

Soon I was standing inside the eLounge, and a sales person walked over to me and asked if I needed any help – in crystal clear voice!  I had yet to go through any voice setup, so I doubted my own voice would be heard, but it came through very clearly – with no hassle whatsoever.  I probed around the build, asking the sales staff about the architecture of the space (which was quite interesting, actually!).  There were at least 6 others there with me, and more were dropping in every few minutes.  This was the day after they launched, so it may not be typical of daily traffic.  Here is a video showing a preview of the space – though I strongly recommend checking it out in person:

Here is another example from MellaniuM – this one showing the Theatre of Pompey and the Titanic:

There are some questions around whether or not the web.alive team can survive the rounds of Nortel layoffs, after they filed for bankruptcy protection last month.   But industry confidence in this platform is strong, so perhaps there is reason to believe it will remain a viable (although you never can tell).

The experience offered by web.alive is comparable to what 3Dxplorer offers, which uses Java (already installed on many computers) to enable a virtual experience inside a browser.  I had similar experiences in both platforms, with a few browser crashes and a plug-in to download (I didn’t have the most current version of Java installed)- but the experiences were equally impressive.  I have also been following the work of Dave Elchoness with great interest for the past few years – since meeting him in Second Life at his impressive VRWorkplace build, and subsequently on to his work with GoWeb3D using 3Dxplorer.  Here is a brief demo of 3Dxplorer:

Unity3D is another player in this arena, offering what might be the most stunning graphics and easiest loading.  It has yet to crash my browser, and only took a few minutes to load.  They do seem focused almost exclusively on games, and game-like environments – but some of the demos like the one seen HERE certainly makes it seem viable as a tool for architectural visualization.  This short video describes a few of the additional features built into the Unity3D platform.

Then there is ExitReality, that claims to bring “the entire web to 3D” by transforming any existing 2D website into 3D.    It appears as though they are focused primarily on offering this 3D website as their core service set, and those environments (to the best of my knowledge) can only be ‘edited’ by everyday users insofar as you can drag and drop content – essentially decorating the space with found or pre-established libraries of objects.  The other part of their offering appears to be customized, developer-driven spaces – like the now famous Carl’s Jr. 3D website.  Here is an example of ExitReality in action:

It doesn’t appear as though they are specifically interested in offering this platform for architectural visualization, but I understand the standards upon which it is built could certainly be applied toward that end.  For example, applications built using the X3D standard can certainly be custom generated, but I still don’t fully understand the workflow, costs or interface between my CAD file and an X3D-based environment.  The good news is that I’ve been told there is an exciting new project coming soon – specifically related to architectural visualization – built on open standards, which I hope to be writing about soon.

So what can we conclude from these examples?

First off, working with platforms like this requires a certain disconnect between the architect and the virtual model.  There are, in almost every case, at least 1 or 2 parties between you and your model, and they each charge some kind of fee for portability beyond hosting.  In many cases, there remains a kind of mystery around the cost and process involved and each project usually requires a new estimate, similar to the way an architectural illustrator works.

This is quite a bit different than the Second Life or openSim based use-cases for architects and designers.  Though it may be far clunkier and more time consuming, the SL environment functions in a way that architects are familiar with – such as sketches, or cardboard study models, where the designer works on it for a while -takes a step back – works on it some more – gets some feedback – modifies it again, rinse repeat.   You can’t ‘sketch’ with these platforms, and you can’t design as you build.

There also isn’t any strong, central, cross-disciplinary and diverse community to speak of.   Not the way we know it in Second Life and in some of the emerging OpenSim-based grids anyway.  Second Life isn’t just a tool – its a place.  In that place, there are hundreds of thousands of real people – from every conceivable profession.  Gathering quality feedback around design ideas, and being able to explore other people’s ideas and share your own has significant value.  Again, I’m not talking about a perfected and polished final model – but a schematic design, or a design concept.  On a moment’s notice, you can teleport in a dozen colleagues from around the world for an on-the-spot critique of an idea – and you can work out criticism and feedback using realtime modeling tools.  You still can’t easily do that anywhere else.

For similar reasons, these platforms aren’t suited as well for the many facets of architectural education either, other than to explore design precedents and existing architectural masterpieces (which I think has huge potential).  But as far as teaching design itself, the elegance of the simple in-world building tools you find in Second Life are ideal for architectural education – requiring a careful examination of the fundamentals of form, space and order.  Realtime modification and collaboration are key to education as well – which you understandably won’t find in these platforms (yet).

There is also something to be said for the browser crashes I experienced in almost every one of these platforms.  They were definitely insignificant, and I got past it quickly – but I tried almost all of them on 2 different computers, and had pretty much the same experience each time.  Please trust that I’ve grown accustomed to crashes and client challenges over the years – but its one thing to have a separate application crash, and another to have your browser crash.  After all these years of wishing SL could run inside a web browser, I’m starting to realize that – unless it can be razor thin and super stable, I don’t want anything tampering with my browser experience!  That’s where my e-mail and blogs are – and a few other tabs I’ve left open for later reading but haven’t bookmarked yet.  Its somehow a lot more disturbing when my browser crashes rather than another application.  Another nice thing is that when SL gets laggy or crashy I can flip back to my browser for a while and read e-mails, etc. while I wait for things to clear up.  When your browser crashes, you’re left with nothing.  Then you restart and realize the same 3D-world tab was auto-saved and trying to restart itself again – bogging down everything else while you wait even longer. This is a petty complaint, I know – but if I have to download 90 mb worth of content and a proprietary plug-in anyway – is it really necessary that it live inside my browser?

I still think there will be certain types of architectural visualization that will be perfectly suited to this kind of technology, but I think it will remain feasible only for more polished presentations of architectural designs – after the design has been finalized and ready for public consumption (for the time being anyway).  As such, this is not necessarily something an architecture firm would want to be continuously porting to as a design development tool.

I also don’t think we can expect to see mass adoption of this technology by the AEC industry, especially architects, given that they tend to be notoriously conservative.  In this tough economic climate that has left the profession gasping for air, with architects being laid off in droves, the profession doesn’t need better visualization tools – it needs to be re-invented.  Architects and designers who are privy to this generation’s new and increasingly open way of practicing are already sharing ideas and collaborating in new and unprecedented ways.   There is a community network within SL and spreading throughout the whole of web 2.0 that is re-molding and re-shaping the future of their profession – whether the profession knows it yet or not.  They aren’t just looking for ways to prototype the buildings they create, they are prototyping the profession itself.

That said, I see browser-based platforms like web.alive, 3Dxplorer, Unity3D and ExitReality as exciting and unprecedented opportunities for visualizing architectural designs from within an easy-to-access web browser application.  However, I don’t really see any of these as as an all-out replacement or competitor to the way architects and designers use Second Life, since it isn’t a valid comparison.   The offerings fit into entirely different phases in the design development timeline.  In its current state, this kind of technology lives at the very end of the design development cycle – at the point where architects must choose the most effective means of unveiling a design concept to their clients – and for clients to unveil a design to the public.  This is generally the province of illustration artists, who would otherwise offer prescriptive 3D illustrations or animations.

Second Life was never designed or intended to be useful as a professional visualization tool for architects, and attempts to shoehorn designs into that platform (one prim at a time) continue to come up short.  Its value becomes most apparent in other facets of professional practice – not just in final design visualization.   We can’t forget that, until now, there were no other options for multi-user, realtime virtual experience of a design concept.  That seems like more of a cause to be grateful than disappointed, imho.  Those who attempted it within SL, did so as a proof of concept when no other option existed – and enjoyed the many benefits it afforded.  Consider Virtual Palomar West as one example:

Like Caleb Booker pointed out in a recent post:

“the hospitals and hotels built in Second Life as proofs of concepts were great and worthwhile. Tangible return on investment in the form of fast prototyping, where designers could solicit large amounts of feedback from a wide range of users, definitely served to improve those projects.”

In sum, I think browser-based visualization platforms have opened 2 exciting new lanes of travel on the bridge between virtual worlds and architecture:  3D model import from CAD, and easy multi-user realtime access to those models within a browser.  However, they did so with a trade-off.  I think architects and designers would do best to leverage the full spectrum of opportunity afforded by virtual environments of all types, and avoid abandoning one set of limitations for another.