Robert Miles Kemp’s talk was always interesting and occasionally spellbinding, most of all when showing the work in responsive robotic structures. His videos of simple blocks self-assembling into what he called “nano-architecture” are quite extraordinary (sometimes eliciting a collective delight similar to that of The Living at Postopolis! NYC). Kemp situated this within a wider context of interactive and informational architecture, centred around his work at Variate Labs and renowned new media deisgn firm Schematic (and his blog, Spatial Robots) described in a consistently interesting talk, covering many of the primary themes in contemporary interface design – and indeed extending the idea of where and what interfaces are.

Kemp starts by stating his interest in “robotics, new control systems, and user interface displays” and suggesting that the “the line between digital and physical is becoming increasingly blurred.” To illustrate the latter, he shows a video of Microsoft’s vision for 2019 (produced by Schematic for Microsoft Research?), talking through the primary themes. These include “information being attached to objects, objects existing beyond physical space, merging information and physical space into one thing, information embedded into actual structures …” and so on. He notes how this “blurs the line between software and control device; how everything is beginning to get jumbled up.”

“We’re working on all of these things”, he says but he “can’t show a lot of the work he’s working on.”
He’s very interested in augmented reality, with its ability to convey a “seamlessness of information. It’s everywhere in your life. All the way through everything, from manufacturing to physical products to home life.” (Aside: seamlessness is a phrase often used without question, as if naturally a positive (not that Kemp was necessarily saying this). This concept has been discussed in HCI for a long time now, with some more interesting propositions around seamful interfaces instead cf. Matthew Chalmers, discussed here.)

Continuing Kemp’s broad themes, he sees that “simulation is the real thing, information is physical (or can be)” and he states that “architecture should be interactive.” (A statement that also requires interrogating, as well as defining carefully, given how many architects have been talking about interactive spaces well outside of contemporary technology for a long time. One person’s definition of interaction is not necessarily another’s definition of interaction. Similarly, as Kemp would well know, much emerging interactive architecture is 'merely' responsive.)

In his forthcoming book, Kemp will run through a “history and future of interactive space”, which should be fascinating, with his last chapter identifying a range of different fields and advances, mainly dealing with new means of control and interaction via robotics and new trends in user interface design. He then talks through a few of these headings.

Starting with the “Robotics” theme, he describes his work with what he calls “nano-architecture” in 2006, and particularly the “shift from humanoid robots to robots as networks or as series or different parts” He shows videos of the USC ‘spider robot’, and the ‘Hot lips robot’ at Cornell. Also, a 3-storey robot, “huge transformable structures built with networked architecture.”

He shows “self-similar robots”, which create “nest-able shapes of waterproof and solid structures” (more Cornell Uni work). These incredible robots rotate within themselves, with “intelligence built into systems – they find each other and rearrange themselves into configuration which enables a structure, which they can move in unison again.” (Leaving aside queries over the definitions of ‘intelligence’, these are quite incredible videos, showing these blocks tentatively exploring each other, coalescing into formations which then become new, larger units with new capabilities, or scaled/nested versions of the previous movements. Extraordinary.)

With “Interactive nano-Architecture”, Kemp talks about building an architecture out of hundreds of thousands of little robots, each the size of a coin (so not that ‘nano’). This is like a kit of parts (broadly redolent of John Frazer’s ‘An Evolutionary Architecture’?) that can come together to build new physical structures. He suggests that this kind of architecture could enable a “house could become multiple houses, or move to a new space altogether.” (Again, extraordinary possibilities; well beyond the (wonderful) Sliding House, towards the ideas of Yona Friedman; closer yet to some genuinely responsive surface ideas Carlo Ratti once showed me.)

With “Nano-Scaled Information Displays”, Kemp is asking whether we can build an interface that’s physical. He shows a beautiful film about car design, wondering “whether there is a new type of physical interface”, and suggesting that “advancements in control technology and interactive language will allow for the development of new kinds of interactive space.”

A current example of this might be “Touch and Multi-touch Interfaces”, also indicating the multi-touch as building block – like the multi-touch cells of multitouch.fi.

At Schematic, Kemp says they work on “a lot of hardcore interface products”. He shows some work for a TV interface based around “Gestural” controls. (This seems slick and considered at this brief glance, but you wonder whether many of the gestures will supplant the remote control interface, just as car interfaces have rarely changed. Certainly as TVs are replaced by PCs with vast online media libraries downloading from the cloud, gestures may be useful in those new media-abundant use-cases, but different gestural moves for volume up/down, or channel change? Not sure. Either way, the latest iPhone Remote app is almost there.)

Kemp then moves on to “Expression, Emotion and Cognitive Control.” He describes how “you can control interfaces with thoughts”. Shows the Emotiv project, though notes that he “hasn’t really started hooking this up to architecture yet”. He talks about real-time communication and real-time information, facial, mood-based and cognitive processing. Then “new hybrid UI controls”, centred on touch, gesture and cognition.

He says “architecture is about designing gesture and interaction – and also information exchanges. More and more these things are becoming this one combined thing. The information and physical object are two things working together in real-time.”

He talks about current user interface trends, such as “real-time flexibility, customisation, personalisation.” He sees “everything is becoming more customised, as long as you have your personal mobile devices.”

He shows us a visual calendar project, by Potion Design Group in New York, which is “state-based and not page-based”. (I think it's this one for RISD; not sure. I like this one though.)

He shows some of his designs for the Nike Beijing 2008 video player for Mexico, in the context of illustrating how architecture informs his work, “developing systems that move in space.“ (Nice piece of work.)

He describes some of these inferences from architecture – the importance of positioning physical objects; using geomterty to organise information. He describes work on the Grey’s Anatomy website, pulling back blog entries about Grey’s episodes, where the organisation correlates date, popularity etc. (Again, a nice piece of work – akin to visualising our ripples ideas five years ago – but I’m personally not sure about how much genuinely ‘architectural’ thinking informs these visual interfaces. A different architecture informs the definition and organisation of the originating data – which can be perceived spatially – yet the visual and conceptual organisation of data is its own discipline, with quite different idioms, models etc.)

He then talks about “digital visualisations of real-world concepts”, such as the much-blogged-at-the-time Current State iPhone app indicating energy consumption in real-time. He describes the “use of space to organise complex real-time information”, with reference to the much-blogged-but-rarely-seen Photosynth.

Kemp then moves onto “tangible visualisations of information”, describing work with Microsoft on a Surface table music application, and then a food-based application. Then the “seamless information display and control for home media network” for the Time Warner Cable Symphoni project, appearing soon. This addresses the problem of aligning an interface across the multiple devices that exist in many homes. (Again, interesting-looking work though difficult to assess from a brief glance of course. The Surface application looks interesting.)

Kemp closes by stating that, now, ‘architecture is digital and physical. It’s more important than ever that architects design this relationship between information and space.”

In the Q&A, I can’t resist picking up on this last statement. In my experience, over two decades of working in and with interface design, I’ve heard this statement many times yet seen few architects with a genuine facility for interface design, interaction design, information architecture. It seems an assumption that is rarely realised. You only have to look at most architects’ websites to see the problems – a cheap shot, but a valid one (also made by Dwell recently) and there many logical reasons underlying this. This may change as some interfaces become more physical in essence (though surely industrial designers have a head-start here?) There's no essential reason why architects would have any more facility with interaction design than any other design discipline. Just as some architects make wonderful furniture designers (especially Italian and Japanese architects), yet many don't. Some architects make wonderful boat designers; many don't. And so on.

Kemp politely disagrees, which is fair enough. He certainly sees the potential in the approach of architecture in terms of interface design. 

I also see the potential – at least when architecture is reconfigured more broadly as concerning “spatial intelligence”, after Leon van Schaik’s recent book. It’s just a real stretch for architects, currently, and will require a radical redrafting of education and practice. To be clear, many architects clearly would make good interaction designers. Some already do (Kemp, for instance). It’s just that most currently don’t. At all. It’s a quite different design discipline, despite the field becoming increasingly 'spatial' (a term which needs picking apart). Again, industrial design is the potential bridging discipline, it seems to me – not least in terms of understanding ongoing iterative processes, systems, user research-led innovation, never mind haptics, IA, interaction models etc. – and it’s interesting that industrial design’s influence on architecture is as strong here in LA as anywhere.

Regine picks up on the terminology, asking about the difference between ‘interactive’ and ‘responsive’. She notes that many of these new architectural experiences are basically responsive, as compared to the genuinely interactive work of, say, Gordon Pask. Kemp replies that “interactivity augments human experience”, which is an interesting and useful starting point, when taken forward. While responsiveness could also do this to a degree, this idea of “augmenting human experience” suggests a deeper, more meaningful connection.

Regine also asks about the current context of projects, of “crisis” – either the GFC or of environmental sustainability. She wonders whether “these projects exist in a world parallel to this world? Can robotics give some answer?”

Kemp replies that he think that robotics could be useful, but that “he’s not 100% interested” as such. He finds “green architecture interesting” but isn’t really motivated by it. Regine finds it odd that, with these projects, there’s “never someone there talking about sustainability?”. Kemp says “Never. I’m not going to be one working on those projects.”

Leaving this aside, Kemp’s talk was a useful, informative and wide-ranging whistle-stop tour of contemporary concerns in physical interfaces and interactive architecture. His book should be worth picking up if it gives the same broad overview. Similarly, his work at Schematic indicates an impressive facility with the medium. However, by far the most compelling examples he showed were the ‘self-similar robots’. Seeing blocks seek each other out and twist around each other, forming new shapes which then have new spatial capabilities, was enthralling. Most people I talked to agreed. It’ll be worth keeping an eye on how Kemp converges the techniques of digital design with these responsive robotics. Fascinating work.