Harvard GSD



Sep '04 - Aug '08

  • anchor

    Hiroshi Ishii on Ambient Interfaces

    By bryan boyer
    Feb 26, '05 1:54 PM EST

    Having waived out of the Digital Media class, I'm taking Toshiko Mori's Innovations in Structures as an elective this semester. The class is set up as a series of guest lectures to increase the dialog between architects and engineers. According to Mori, this class is, in some respects, a continuation of the work from Immaterial/Ultramaterial. This week was the first guest lecture presented as a discussion about "life at the border of bits and atoms" by Hiroshi Ishii. Ishii's Tangible Media group at the MIT Media Lab has been working on exciting projects. Sensetable, Topobo, I/O Brush, and SandScape are just a few amongst many that you may have seen before.

    While Ishii covered a number of topics, most relevant to the audience here was his brief discussion of architectural space as interface. In all of his work Ishii is interested in the convergence of input space and output space. In architecture this becomes manifest in projects like I/O Bulb that propose a architecture-scale application of the sensetable technology to effectively add a parallel layer of computational actionability to normal physical actions. This sort of thinking points to interesting possibilities but ultimately one wonders about the viability of large scale, centralized, environmental computing. Even Ishii's own work tends towards an object scale, focusing instead on multiple and task-specific projects.

    Ishii implicitly acknowledges this in the specifics of the architectural projects he showed, primarily Ambient Fixtures. What I find compelling about the project is its focus on the periphery, ambient experience of information as a counterpoint to the dominant, center-focused interaction we're used to. It's not surprise that Ambient Devices is a spin off from Ishii's group at the Media Lab, given that they are beginning to implement some of the ambient informational technologies that Ishii is concerned with.

    With architects more and more considering information and various media as an important part of their buildings, Ishii's five points about the ambient interface of architectural space are good to think about.

    • Browserless Information should be glancable and require no navigation
    • Calm Should be seamless with the environment
    • Persistent connection Information must be current, and regularly updatable
    • Decision driven data Should be personalized and summarized to help users make decisions quickly and easily. "Should I bring an umbrella with me today?"
    • Private Information should be encrypted for privacy

    Image from the Design Boom.

    The first hints of these ideas are just starting to emerge in architecture with projects like UN Studios' recently completed Galleria in Seoul, South Korea and Diller+Scoffidio's Blur building from 2002. While neither project addresses information at the personal scale, they respond more broadly to environmental factors. Indeed, as a building made of mist, Blur relied on wind and barometric sensors to help keep its cloud localized over a steel skeleton. Sensors communicated environmental conditions to a central computer which then adjusted the rate of flow to an array of nozzles covering the skeleton allowing for real-time compensation for strong winds and other disruptive conditions. In effect, the building took advantage of ambient sensing to make constant decisions defending itself from the environment.

    Image from the Archinect image gallery

    If Blur is using ambient technology to act in its own interest, UN Studio's Galleria operates solely for the entertainment and attraction of the public. Currently the building records the day's weather and replays this footage as a slow motion sunset depicted via a facade of 5000 frosted glass discs backed with LEDs. Ishii mentioned that the Hancock Tower in Boston displays a prediction of the next day's weather based on the pattern of lights at night, but the Galleria can go one step further. Displaying the previous day's weather is pure entertainment, but with enough data built up the building could begin to use previously recorded weather videos as a premonition of what's to come. Based on forecasts, the facade's computer could search though its database of recorded days and find a past day that closely matched the forecast. Animating the facade with this video content would then give passersby an idea of what tomorrow might be like, instead of what has already passed today.

    What's most promising about Ishii's conversation is that there may be a future without video projections on loop or random LCD screens regarded as avant-garde. Implicit in all of his work is the notion that static computing is boring. Computational power, storage space, and bandwidth are reaching a point of abundance that makes invention easier and, importantly, more realizable. If we're currently living life at the border of bits and atoms then I look forward to a future of dual citizenship.


    • dillup.

      There's a project in the latest Verb : Matters in Stuttgart, I believe, where they used some kind of computerized water patterns to communicate conditions to the city. It was some kind of civic building... anyway, sounds related to what you're talking about.

      Feb 26, 05 2:15 pm  · 
      bryan boyer

      dillup- Actually, I was trying to track that project down, but couldn't remember the name of the architect and my copy of Verb is at home. But yes, it is relevant. Though I'm not sure if the information that building is dealing with is live in any way or if it's just replaying pre-programmed actions.

      Feb 26, 05 2:18 pm  · 
      juan moment

      If you haven't come across it already, Information Arts by Stephen Wilson is a great book in this area. I am not sure, but I think the project you are trying to track down might be the interactive facade at the Zeilgallery in Frankfurt by Christian Moeller.

      Kinetic Light Sculpture

      Feb 26, 05 3:38 pm  · 
      juan moment

      sorry - here is the link: Kinetic Light Sculpture

      Feb 26, 05 3:49 pm  · 
      Mason White

      bryan and dillup -
      maybe you mean the jurgen mayer h stuttgart town hall rain facade called "pitterpatterns" ... and yeah i think you are right, it is pre-programmed patterns. kinda wallpapery.

      oh and actually,klein dytham have done some intereseting work with interactive surfaces (some digtial some analog) like this (digital) one:

      and this (analog) one:

      Feb 26, 05 4:16 pm  · 
      bryan boyer

      mason wins the prize.

      Feb 26, 05 4:23 pm  · 

      decoi's aegis hyposurface is related.

      Feb 26, 05 7:03 pm  · 

      how about this one, the BIX facade of the Kunsthaus Graz. It's not going to look obsolete in five years time because it's already so retro:

      This field (particularly the "calm" and "browserless" aspects) is often called "ubiquitous computing", although Ishii doesn't seem to use that phrase.

      Feb 27, 05 9:21 pm  · 
      bryan boyer

      theo- ubicomp is generally considered a broader field than what Ishii is interested in. His work specifically targets the methods of interface (hence the name of the group: Tangible Media) and takes for granted a lot of infrastructural issues like ubiquitous computing. I don't remember him using the term, but it was only one brief presentation. Actually, if we're trying to use the buzzword "ubiquitous" to describe his work I think the phrase "ubiquitous interface" may be more appropriate. I'll ask him what he thinks next tuesday.

      I pointed out the Galleria and especially the Blur building because, in a basic way, the buildings themselves are empowered to act upon the data they're collecting. In contrast, Kunsthaus Graz, the hypersurface, and even pitterpattern are all entertaining but none of them are much more than specialized display surfaces. That is, there's no specific system at work-- just a building-sized browser.

      Feb 28, 05 12:19 am  · 
      Derek Lindner

      Eddie Opara (atg, yale, 2x4, imaginary forces) is teaching a class on Dynamic Skins this semester here at Columbia--haven't seen the projects yet, but given his interests and his design for the facade of 745 7th Ave in Times Sq, it ought to produce a number of interesting ideas in this area. I'll try to post something in a few weeks, when there should be some projects to see.

      Feb 28, 05 12:48 am  · 
      George Showman

      Excellent, informative post and comments! Thanks to all involved. Brian's clear distinction between "browser" buildings and buildings that actually act on their data is nice, though I'm not sure how that Galleria building is anything other than a big, bright browser (but I don't know the project).

      One thought, though: don't you think that a lot of this merging between physical architecture and electronic data/cyberspace/etc. will actually occur somewhere between the user's eye and the object? I.e. won't we soon have the headset technology to overmap the physical world with layer upon layer of data through a small retinal projector? The best example is perhaps Steve Mann, formerly of MIT, now at the university of Toronto. Here's a link to the most relevant work he's done:

      (check out the .avi's, although they're frustratingly short)

      And even if you're not into the "everybody has a headset" vision, wireless devices are reaching the point where everybody is taking pictures of everything around him. The next step will be to overlay information on those pictures. Steve Mann is basically walking around a city covered with ornament he himself is generating on the fly.

      Seems a lot more reasonable than trying to turn buildings into projectors. But I haven't thought all of this through yet.

      Mar 6, 05 10:38 pm  · 
      bryan boyer

      george- although turning everyone into a cyborg does sound like fun, I think one large benefit of room or building-scale technology is that it facilitates group interactions mediated by technology. For instance, the set of sensetable projects from the Media Lab are great collaborative aids when a group of people are working through an idea.

      Mar 6, 05 11:06 pm  · 

      Block this user

      Are you sure you want to block this user and hide all related comments throughout the site?


      This is your first comment on Archinect. Your comment will be visible once approved.

    • Back to Entry List...
  • ×Search in:

Affiliated with:

Authored by:

  • bryan boyer

Other blogs affiliated with Harvard University:

Recent Entries