If you are at all interested in the world of computational design, please please read this blog post by Daniel Davis on his blog, Digital Morphogenesis. It is a great survey of blind spots and false assumptions within the CD community, told through a minibiography of William Mitchell.

I’ll also be posting in a week or so a short interview with Daniel, alongside a lot of other academics and practitioners, discussing workflow and intellectual property issues in digital architectural practice.

Presented my final project for CS50 (basically Programming 101) at the “CS50 Fair” last Friday. I got quite a good response – some people actually brought their friends back to look at it a second time. There weren’t a lot of games at the fair (Harvard undergrads seem to be a very pragmatic lot), so it probably seemed fun by comparison.

My project involved attempting to use a rudimentary optical flow algorithm as the main control for an interface. I chose a game as, well, it’s more fun to talk about than anything else. I chose to re-imagine the old “Missile Defense” Atari 2600 cartridge, although in retrospect something less violent probably would have had broader appeal. Optical flow is a kind of machine vision that looks at implied motion in a video image – basically, wave your arms or hands and the game responds. I built the game in Processing, an open-source visualization language that I can’t say enough good things about.

If you want to play (and have a computer with a webcam), it’s hosted here as a Java app. Since it’s self-signed you’ll have to tell Java that it’s save to run by clicking “no” on the popup that appears. If you have a PC you’ll probably have to have VDIG 1.0.1 installed to give it access to your webcam.

The best (and also most embarassing) moment at the fair was when I spent about 10 minutes presenting to three or four people whom I assumed were parents or professors. Afterwards I had someone come by and ask me what it was like presenting my project to Drew Faust. Oh, well.

I promise I’ll show something physical one of these days. This video is a modification of the sobel edge detection method below, except for this time I’m taking advantage of the fact that the algorithm actually knows the edge direction to produce vector normals whose length corresponds to the strength of the vector. I’ve also added some ability to change the vector length multiplier and threshold so you can get different effects. This is katy’s favorite thus far.

Next I’m going to try to add frame differencing so that I can get it to do this only when sensing motion…

A quick preview of some of the stuff we’re working on in Augmented…

**UPDATE** oops, forgot about this video as well. This is a first attempt at edge detection (Sobel algorithm) in processing on video files. Could use better noise reduction and runs a bit slow but hopefully I can optimize it easily…

I had a guest lecturer at a recent class who spent an hour showing project after project featuring fantastic custom tools their office had generated, from BIM models connected to a proforma through Excel, to LEED client websites, to urban-design Revit templates and family libraries. When I asked the presenter his stance on intellectual property issues, he said something like “well, naturally we’re all about open source, but we haven’t really thought about how we might share this content.”

Architects need to not only support the idea of open source practice, but be actively involved in its promotion and dissemination. Here, I’ll bullet point it for you:

1. Digital practice is becoming ubiquitous.

2. Architects are increasingly required to navigate between multiple software packages / modes of working in order to complete a project.

3. To get this done, architects are increasingly required to generate their own digital content in the form of scripts, templates, plugins, spreadsheets, tutorials, etc.

4. Few practices give critical attention to either their workflow or their methodologies to design with new digital tools. As a result, they are not only losing on efficiency but they are allowing the tools themselves to have a much greater impact on design than needs to be the case (i.e. the maximum flexibility of each tool is not being realized.

5. In order to promote tool and method sharing, and to promote a more critical evaluation of design tools, open source practices need to not only be accepted in the architectural community, but actively promoted and managed.

6. Centralized models to disseminate theoretical or practical information already exist (archinect, design reform, evolutionzine, Rhino plugin websites, etc.) What is necessary more than ever is the sharing of methodologies and ideas on how the siphoning of information between various programs can lead to meaningful, beautiful, solid design.
—–
Instead of ignoring the power of digital processes to unwittingly influence our work, it is better to embrace the fact that design is now immersed in this virtual medium, and collectively experiment ways on improving, leveraging, and transcending the spatiodigital morass that is contemporary practice.

I had some big plans for this blog this semester. Unfortunately they were somehow railroaded underneath an incredible pile of coursework and group projects. Here is a class-by-class update, maybe with a few images if I can drag them up:

New Frameworks: this class has really demystified the business context that surrounds architecture firms. It’s also reassuring in that Paul Nakazawa can’t go fifteen minutes without explaining how nearly every small or medium size firm on the planet could be run better. Apparently as a collective, we are not business geniuses. My favorite moment thus far was a guest lecture by Tim Love from Utile, followed by a nice discussion on intellectual property, local practice, and how connections generate work.

CS50: this one is keeping me really busy. No super-exciting classes or projects yet, but I am well on my way to really understanding what makes C/C++ tick, which also means knowing a lot more about what’s going on under the hood of my computer in general. Right now we’re playing around with recovering image files, which is kicking my butt.

CAD/CAM: had a team presentation last week – our project was a chair modeled in Digital Project that would automatically adjust its ergonomics based upon your pants size and a few other measurements or preferences, then automatically output the curves needed to mill the chair from a 4×4 sheet of ply. We’re making a second version of the chair as a tiny couch for Paul and Amelia, pics of that for sure when it’s complete.

Augmented Environments: my current team project here involves creating a responsive lighting installation, using Arduino, Grasshopper, and some basic kinematics. We’ve been looking at a Hobermann toy, the SwitchPitch, for some ideas on ways things could move. More detail on this one as it develops. I’m learning a lot of Processing, so I might also have some fun video toys to show off soon.Computational Design: this has been half C#, half vector math. We’re just now getting into a project that will involve using computational methods to design a canopy. Once again, more soon as things develop, for now some subdivision and xform eye-candy.

Thesis: what, are you going to bug me about this too? Jeez.

As the program I’m in can seem kind of nebulous, the best way to describe what I’m doing is to show what I’m taking. This semester is largely turning into a kind architectural technology survey, with a little alternative practice thrown in. To wit:

CAD/CAM, taught by Martin Bechthold – tooling/cutting/parametrics/robots. This year the course has a focus on preformance, meaning some sort of real metric that’s controlling operations. Like most of my classes this semester, focus is largely on practical application but it looks like we might be given a little leeway for flights of fancy if we angle it the right way (we are providing the definition for what kind of “performance” we are after).

What I’m looking forward to: learning the real basics of G-Code and robotic control languages, and getting a little deeper into Digital Project. Also, using a water jet on a robotic arm is pretty cool by itself.

Computational Design, taught by Panagiotis Michalatos – form finding, problem solving and visualization for physical and spatial problems. A lot of finite element analysis and very complex geometries, with some interesting physics and visualization techniques largely pulled from gaming environment. The professor has written a great deal of his own software for the work, and while quiet appears to have a pretty good sense of humor. His work can be found at sawapan.eu.

What I’m looking forward to: actually performing the kinds of structural form-finding that usually gets relegated to consultants.

Augmented Environments, taught by Mariana Ibanez and Allen Sayegh – sensors and actuators, responsive environments. A very Media Lab sort of course (in fact two people from the ML are cross-registering). Is going to be a lot of Rhino, Grasshopper, and Arduino, which are all things I have very little knowledge about. Might be an uphill climb.

What I’m looking forward to: building a little robot. Duh.

Intro to Computer Science, taught by David Jay Malan – this is actually an undergraduate class I’m “upgrading” to graduate status. Harvard has been smart enough to not only make Programming 101 a valid core curriculum option, but also to allow the professor to give it a focus on contemporary solutions. The final project can be a mobile application, Web 2.0 project, game, etc. While it’s a little grating to be one of the only people in the room that wasn’t born in the 90’s, I’m already getting immediately useful information out of this class.

What I’m looking forward to: actually being able to back up my bullshit with knowledge when it comes to APIs and C#.

A New Framework for Practice, taught by Paul Nakazawa – my sole non-tech course this semsester, focusing on alternative practice in architecture and the changing post-recession (or mid-recession, depending on your outlook) architectural business environment. Lots of round table environments with someone who has been around the block more than a few times.

 What I’m looking forward to – getting some straight dope on how some well-known practices stay afloat.

That’s the list. I’m hoping that getting a lot of this stuff out of the way this fall will allow me to go a little wider in the spring. We’ll see how it works out.

After marathon ikea shopping, some craigslist adventures, and the delivery of two very happy toddlers, we are on the ground and running in chez Regnier, Cambridge edition. School doesn’t officially start until the middle of next week, but until I have real information to share here are some quick impressions.

-Two semesters is no time at all… I’m having to craft my entire year at one go.
-Everybody I’ve talked to so far in my program (M.DesS) has some skill or knowledge that I wished I had. I’m hoping that the format of the program is such that things will end up very collaborative, as I’m planning on picking a lot of brains.
-Harvard spends a lot of money on entertainment – oodles of lectures, mixing events, symposia, etc. It’s kind of continuous architectural feast that moves so quickly there’s no time for toe dipping – you have to just jump in.

On the non-school side, our neighborhood is beautiful and chock-full of children, parks, trees, and charm. Apparently we fell into the middle of the Park Slope of Boston. People here seem much more likely to do their own gardening and child rearing here than in LA (good), but are somehow even preachier about their chosen methods for Saving The Planet (when we arrived I parked in front of a Swedish vehicle with a bumper sticker that read “I’d Rather Be Organic Gardening”).

More about all of this the next time I come up for air.

The future would be a lot better off if all of us – interaction designers, architects, technologists, journalists – would agree never to use the phrase “augmented reality” ever again. Few two-word combinations make me more uncomfortable – each time I hear it I squirm in my seat, holding back the desire to inform others against their will.

A short thought experiment: you have a device that, when you look through it, makes the sign on a restaurant you are looking for appear to flash on and off, to help you recognize it walking by. Is this augmented reality? Ok, you have another device – this one actually triggers a switch in the building itself, so that the sign lights actually do flash on and off, when you walk past. Also augmented reality? OK, a third device, which is actually just a friend with a phone, calls the restaurant, describes what you’re wearing, and somebody sits by the window, and when they see you run to the switch and – you guessed it – flash the sign lights on and off.

Augmented reality is often described as a layering of information on top of what we normally perceive. This draws an entirely unnecessary distinction between data and physical space. Aren’t street signs technically augmenting your knowledge of the city? The fact that you don’t need a special phone to decode them doesn’t mean they don’t serve essentially the same purpose as a GPS. Your brain is already a fantastic device for making connections between places and data, not to mention pulling data directly out of context. The complex interaction between people, site, and data already has a fantastic name – reality.

We’re putting blinders on our (augmented) vision by committing ourselves to this idea. As long as people keep talking about augmented reality, research and thinking about wayfinding, interaction, and mobile computing is going to unnecessarily complicated by ideas of visual layering, holograms, and funny headsets that nobody in their right mind is going to wear out of doors (particularly when it is raining). In the book “Makers” by Cory Doctorow, the protagonists invent a device, shaped like a garden gnome, that uses gait-recognition to identify certain people walking past and give them a simple message- “don’t forget to pick up the dry cleaning,” perhaps. There you go- your reality has been augmented. No gps, 4g, or silly glasses required.

It seems obvious to me that as our interaction with data becomes more sophisticated, the interface is more likely to recede than present itself. Auditory or even tactile cues are more efficient, subtle, and robust than anything that needs. Even tech that does show a picture is more likely to use the wall next to you (lasers, projectors) than some complicated extra piece of headwear. And finally, if technology is getting smarter, isn’t it more likely that it can be left alone for longer periods of time, and we’ll be spending less time interacting with it?

Once you replace this idea of augmentation with the observation that reality is simply getting smarter, faster, and more complex, it’s much easier to see where things are going and what we might improve. Let’s hear it for reality. There is nothing more, no matter what we add.

The ensuing melee in the comments section of the Patrik Schumacher parametricism follow-up makes for decent (albeit nerdy) entertainment. While it isn’t necessarily informative, it does provide the following telling response from Patrik himself:

….to make the claim that an epochal style is in the making requires me to define the challenge that compares, in terms of urgency and universality, to the industrial revolution which spawned Modernsim. …The stable, homogenized society of fordist mass production has given way to the dynamic, multi-cultual network society of today. (Contributing factors: globalisation, micro-electronic revolution a.o).

This confirms what I had suspected before – that one of the goals here is to cast Parametricism as a global, hegemonic style, grown organically out of a cultural and technological zeitgeist. This is fundamentally wrongheaded, for the reasons (beyond the usual it-will-leak claptrap) I will outline below.

1. Post-Fordistnonlinear. This is the most basic and often repeated argument. Representing a “dynamic, multi-cultural network society of today” so literally with t-splines and doubly curved surfaces, is a somewhat simplistic way to claim the expression of the zeitgeist. It smacks of the dumbed-down, literalist “reading” of deconstructionist authors a decade ago. For something so multifaceted and dynamic, this methodology is awfully simplistic and rigid.

2. Hearts and minds. As many a kitchen appliance ad will attest, modernism tapped into a popular, collective need for order and control in the postwar western world. This providential dovetail between academic thought and popular opinion helped push a particular modern mindset towards dominance over other contenders for the title. I might be wrong, but I have seen nothing to suggest any kind of popular will that might lend a hand towards a hegemonic Parametricism.

3. Bean Counters. Modernism had, as a rearguard, arguments for efficiency –and, well, modernity– that provided a rational aura that played well with planners and CEOs. Parametricism has… the Bilbao effect?

4. $$$$. Modernism had a symbiotic relationship with concurrent factory production that made it easy to disseminate and reproduce (often with dubious quality). In a sense modernist construction methods have only reached maturity in the last 20 years, after the declared death of the movement. Parametricism is unlikely to connect naturally with industrial capabilities, except those associated with expensive consumer goods (furniture and automobiles, to be precise.) It will take a ground-up rethinking of our methods of production to make this kind of architecture within the realm of possibility for 99.9% of projects. Industrial revolutions tend to be led by industrialists, not by artists. See #3 above.

5. There won’t be another Beatles, either. Modernism’s brief mid-century hegemony was an anomoly – an architectural monoculture in an age of monocultures. If every other artistic field is currently undergoing fragmentation and localization, what chance does architecture have to unify under a common flag, particularly one lacking emotional and practical ties to the majority, as evidenced above?

Even if the aim is not to dominate but simply gather forces as one style among many, the chances of Parametricism evolving beyond boutique projects with high budgets and “enlightened” clients seem slim, even as other parametric methods are likely to come to the fore in the near future. The more likely outcome is that parametrics will become an integral part of design and construction – occasionally with Parametricist stylistic effect – but under a different name with different end goals. It is equivalent to Gordon Bunshaft writing a manifest for “Curtainwallism,” or perhaps more accuratly equating Le Corbusier’s Five Points with the modernist project as a whole. Perhaps Patrik is doing everyone a favor by lobbing the inital volley in a game that will be played out in the next half century; in any case I am afraid that this movement may do more harm than good, by constraining the future to a single stylistic interpretation.