KINETIC, RESPONSIVE AND ADAPTIVE: A COMPLEXADAPTIVE APPROACH TO SMART ARCHITECTURE
Presented at and published in the proceedings of SIGRADI 2005 international conference, Lima, Peru.
Mahesh Senagala
College of Architecture
University of Texas at San Antonio
501 W Durango Blvd
San Antonio, TX 78207
USA
[email protected]
Abstract
Smart architecture is fast becoming a buzzword in architecture and related disciplines. However, it is not
entirely clear what constitutes smart architecture and how relates to or differs from such closely related
camps as responsive architecture, performative architecture, kinetic architecture, and adaptive architecture.
This paper poses the essential and critical questions about smart architecture from a complex-adaptive
systems point of view. The paper also illustrates the attributes of smart architecture with a number of
seemingly disparate, yet conceptually connected design developments.
1. Intrdoduction
Smart architecture is fast becoming a buzzword in architecture and related fields. Although AT&T toyed
with “intelligent buildings” back in 1982. Graham and Marvin, 1996), much of the research and
development work in this important area is still in its infancy. Although significant amount of work has
been done in the area of “smart houses,” scalability, portability, and conceptual clarity have been the
limiting factors in extending that research to larger building applications. Trulove, 2002).
If we scan the literature on the subject, It is not entirely clear what exactly constitutes smart architecture
and how it relates to a number of closely associated camps such as responsive architecture, performative
architecture, kinetic architecture and adaptive environments. What does it mean for architecture to be
smart? How smart is too smart? How is smartness measured? How does it differ from previous attempts at
making architecture mechanically and/or computationally intelligent? How does it relate to or embrace
design computing, computer aided architectural design and energy efficient design? Is there a possible
computational and architectural framework that can be used to provide the necessary direction to the
Page 1 of 13
myriad academic and professional pursuits currently underway at various institutions? A number of
pursuits -- previously purely technical or purely theoretical or predominantly environmental camps -- have
merged and begun to share their concerns under the rubric of smart architecture. The paper attempts to
outline a complex-adaptive systems framework for smart architecture, outline an agenda, and connect the
dots of some significant conceptual, technological and architectural developments in this direction.
2. A Complexity Framework for Smart Architecture
One of the goals of this paper is to begin to define what legitimately constitutes smart architecture. This
task involves defining the attributes of smartness. The notion of smart architecture is wrought with
unintended vagueness. Vagueness can be good sometimes to enable multiple and competing notions to
develop. However, there has been no fundamental discourse about what constitutes smart architecture and
what its principles are. Is anything with a microchip smart? Is anything networked worthy of attribution of
intelligence? What are the benchmarks for smartness in architecture?
Such a definition might lead to, a qualitatively and quantitatively measurable quotient, which the author
likes to call “smartness quotient” or SQ. But before we can attempt to measure a building’s SQ, we need an
understanding of the attributes of smartness. As a systems theorist with a particular interest in complexadaptive systems, I find a number of basic systems concepts to be of great help in accomplishing the
present task of framing smart architecture.
2.1. Complicated versus complex
Complexity is not an oft talked about concept in architecture. Robert Venturi’s Complexity and
Contradiction in Architecture does venture to some extent in connecting with the notions of complexity in
other disciplines while focusing almost exclusively on mannerist complexity of meaning, surface,
iconography and historic references.
What is complexity? The word complexity is a derivative of the latin root complexus, which means totality
and embrace. Complexity refers to a systemic totality and the interrelationships between various
subsystems. This differs from common parlance that complexity is opposed to simplicity. The word
Page 2 of 13
simplicity descends from latin roots sim+plec, which literally means single fold. The true opposite of
simplicity is complication, which means to fold together. While there is a synonymous relationship
between complexity and complication, they are distinctly different notions. Complex systems can be simple
or complicated. Complicated systems are not necessarily complex. Complex systems, as they are
understood today and defined in no fewer than millions of words of cross-disciplinary discourse, are
closely related to such phenomena as adaptive systems, non-linear systems and living systems. In complex
systems there is an element of learning and adaptation. There is also an element of self-awareness, which
differs significantly from automation. Paul Cilliers, a leading proponent of complexity theory sums it up
thus:
The concept ‘complexity’ is not univocal either. Firstly it is useful to distinguish between the
notions ‘complex’ and ‘complicated’. If a system – despite the fact that it may consist of a huge
number of components – can be given a complete description in terms of its individual
constituents, such a system is merely complicated. Things like jumbo jets or computers are
complicated. In a complex system, on the other hand, the interaction among constituents of the
system, and the interaction between the system and its environment, are of such a nature that the
system as a whole cannot be fully understood simply by analyzing its components. Moreover,
these relationships are not fixed, but shift and change, often as a result of self-organization. This
can result in novel features, usually referred to in terms of emergent properties. The brain, natural
language and social systems are complex. Cilliers, 1998).
Truly smart architecture has to be complex, though not necessarily complicated. This and other systems
concepts will be elaborated further in the following sections.
2.2. Automatic versus autopoietic
Smart architecture is often confused with automated architecture. Smart systems are confused with
automated systems. While smart systems do involve great degree of automation they go beyond mere
automation to embrace complex cybernetic processes and learned behaviors.
Page 3 of 13
Autopoietic, on the other hand, is about self-production or self-organization. Organisms, corporations,
societies, and minds are self-producing and self-organizing. Also, as Maturana and Varela define it,
autopoietic systems have a clear systemic boundary and closure that makes them autonomous. Smart
architecture can be truly smart only when it is truly autopoietic. Mere automation and mere use of logical
circuits and discrete sensory mechanisms are not sufficient for a system to be smart.
One example of an autopoietic architectural system is the so-called Topotransegrity project.
Figure 1: Topotransegrity by 5 Subzero
The Latent Utopias exhibition curated by Zaha Hadid and Patrick Schumacher featured a groundbreaking
kinetic responsive prototype named Topotransegrity. It was a curious mix of topological manipulation
though pneumatic spaceframe structure. The project, designed by 5Subzero, a budding group of architects
from London, featured surfaces that can be manipulated through either an automated control mechanism or
a real-time feedback system or a pre-programmable system. The system, if developed as illustrated here,
could become a self-organizing system that approaches a spatio-temporal smartness with a high SQ.
Figure 2: Topotransegrity Prototype in Latent Utopias Exhibition, Graz, 2002
Page 4 of 13
The system consists of a kinetic spaceframe driven by three sets of Festo pneumatic pistons. The original
prototype was designed to sense the audience movements through pressure sensitive mats and translate the
impulses into valve operations controlled via the computer. The scenario depicted in Figure 4 envisages
some dramatic architectural possibilities. Topotransegrity proposes to program the walls and floors along
with people and events in an all enthralling four-dimensional framework for smart architecture.
Figure 3: Pneumatic Pistons
Figure 4: Topotransegrity Scenario
Figures 5, 6 and 7: The Muscle by Oosterhuis_Lénárd
More and more architects are beginning to explore the form and format of such smart and supple
architecture. The Muscle, a prototype featured at the Non-standard Architecture exhibition held at Centre
Georges Pompidou from December 10th, 2003-March 1st, 2004. The Muscle was designed by the Dutch
design firm Oosterhuis_Lénárd and dons a pneumatic structure that behaves like a gigantic, digitally
mediated muscle. The building would flex, contract, expand and mold itself to suit changing programmatic
conditions over time and in real-time. The architects propose that the place be used for a variety of
activities such as a disco or a television studio or a meeting place. The synthetic muscles of the Muscle
react as people move near the sensor points. Another way to manipulate the structure is by moving the
sliders on a remote computer screen. Thus, the building becomes spatially interactive and can be plugged
into the Internet. This alien-looking blob is not necessarily how buildings might look as a whole in the
future. Some buildings might look that way. However, the Muscle, as ONL proposes, can literally be used
Page 5 of 13
as a series of muscles in a building to control any environmental or other parameters in real-time. Perhaps
the buildings might be able to literally express their feelings by flexing their facial and spatial muscles,
thereby leading to truly complex, kinetic, performative, and supple architecture.
2.3. IP versus XP: A Building is Not a Box
A building is a network for living in. While all autopoietic, complex-adaptive systems have a clear sense of
autonomy, they are a clear part of a larger system. Just as a human individual has a clear-cut identity, every
human being in an integral part of a large life-world system. Likewise, smart systems would be very dumb
if conceived as self-contained systems largely disconnected from the larger world systems. A network is
always a more capable, adaptive and complex system than any of its components. Just as a computer as a
box that is not connected to the larger networks is very limited in its role in our societies, architecture
conceived as boxes that are not a part of larger networks is also very limited in its smartness.
A network is an interconnected system with a certain structure of relationships. Kevin Kelly, executive
editor of WIRED magazine summed up the essence of the current paradigmatic shift: “the central act of
coming era is to connect everything to everything. All matter, big and small, will be linked into vast webs
of networks at many levels. Without grand meshes there is no life, intelligence, and evolution; with
networks there are all of these and more”. Kelly, 1994). Nearly five decades ago, long before the computer
or the Internet became popular, Teilhard de Chardin prophetically proclaimed that the human evolution is
heading toward a global coalition of an interconnected world. de Chardin, 1961). He called such a world
noosphere. the sphere of interconnected human consciousness).
Intel’s new WiMAX technology buoys these possibilities by preparing to introduce cell phone-like
ubiquitous and wide-ranging network coverage for laptops and other computational devices.
www.intel.com/netcomms/technologies/wimax). A WiMAX-equipped laptop or computational device can
stay connected to the Internet all the time without any wires. Such a technological network would be a
crucial turning point in the journey toward the emerging noosphere.
At a time when even the toasters and refrigerators are being hooked up to the Internet, architecture is not
going to be left too far behind, despite the best efforts to resist the evolution by the profession’s
conservative core. Connection is the keyword, the buzzword and the overarching concept of how buildings
Page 6 of 13
could become networked computer systems. General Motors’ OnStar ® vehicle security system has already
transformed our automobiles, which are now controlled by dozens of embedded computers and networked
via satellites, into GPS-powered real-time network nodes. An OnStar operator can access most of the
critical systems of an automobile remotely and suggest or coordinate a course of action at the touch of a
remote button. The biggest evolutionary jump for automobiles is not in their growing engine size or
seductive body shape; it is in the pervasive computerization and wireless digital networking. The mobile
space of the automobile has been transformed into an interactive real-time network node capable of keeping
us connected to the rest of the world. Automobiles are already a part of the emerging noosphere.
Architecture is also becoming a part of the post-spatial network ecosystem. The current preoccupation of
the profession with digital fabrication. based on manufacturing paradigms) and complexly curved surfaces.
visual complexity) are both based on outmoded or soon to be passé models in author’s opinion based on the
rapid evolution of smart networks and their market share in the coalescing global economy. Architecture
would be better served to take a systems approach, particularly the complex-adaptive approach of IP.
network) in contradistinction to elitist, formal and XP. box) based approach.
.
Figure 8: OnStar At Home® Architecture. Courtesy Internet Home Alliance.
The Internet Home Alliance. internethomealliance.com), a remarkable cross-industry collaboration
between GM’s OnStar, Invensys, ADT Security Systems, HP, Panasonic, and many other corporate
partners have launched, in early 2002, a post-spatial initiative to integrate OnStar’s Virtual Advisor®
service with home security control, telecommunications control, and climatic control from any Internet
enabled appliance anywhere in the world. This system also gives the customer visual access to his or her
house at any time. Garage doors, access doors, windows, all the major home appliances, HVAC system,
security system, and telecommunication systems are networked using the HP Application Server 8.0
Page 7 of 13
framework as gateway. The participating customer would be able to access any and all of these aspects in
real-time from any computational device such as a cell phone or a PDA or a laptop. The customer would be
able to remotely turn on or off the appliances such as kitchen stove or refrigerator. Panasonic’s smart
doorbell would notify the customer anywhere, through audiovisual access, anytime his or her doorbell is
rung. Thus, appliances, HVAC systems, security systems, and a host of other building systems are
computationally networked and connected to the global nomads – the home owners.
2.4. Plug and Play Architecture, a Building Operating System
There need to be a consistent set of standards for various component of smart architecture to be aware of
their role and relationship to the larger whole. There also need to be a set of standards for various
architectural components to follow the computational model of plug-and-play. Perhaps the Bluetooth model
is also an apt one. When a window is installed in a wall, the window needs to both be aware of its location
and purpose, as well as its relationship to all other components in its subsystem. A floor tile needs to know
its spatial, temporal and functional relationship to the rest of the components. All these tasks could be
accomplished by, what the author likes to call, a Building Operating System. Only then can we truly evoke
the notion of smart architecture.
The Responsive Environments group at MIT Media Lab has devised a sensate floor system called Magic
Carpet. It cannot fly, but the magic carpet does know a lot about who is on it. In this system, a series of
piezoelectric wires in X and Y directions are used to sense the footstep dynamics such as pressure and
movement. The sensing medium used here is an inexpensive shielded wire that is capable of producing a
small voltage. around 15 Volts) spike when pressed. These signals are then processed using various
filtering and clustering algorithms to decipher the location and pressure information.
Page 8 of 13
Figures 9 and 10: Z-Tiles by Responsive Environments Group, MIT Media Lab
The Magic Carpet system has led the team to develop networkable plug-and-play floor tiles called Z-Tiles.
In this system, a series of interlocking tiles, equipped with embedded processors, form an adhoc network
and communicate the prexel. pressure pixel) information to the main computer for further programmable
action. Thus, with this relatively inexpensive and robust flooring system, it is possible to transform any
floor into a sensing surface capable of forming an intelligent environment. The possibilities that this
invention opens up are immense. It is possible to control such architectural characteristics as lighting levels,
location, ventilation, security and other environmental parameters. It is possible to estimate the number of
people and the kind of activity through analyzing the data in real-time, which can increase or decrease the
heating and cooling levels in a room or control the room illumination or activate the entertainment system
or open up the walls or lead to many other architectural possibilities. Perhaps a walk on the floor during
October could activate the sound of the rustle of leaves. Perhaps a dance on the floor could adjust the
acoustic characteristics of the walls to better suit the music. Thus, the very meaning of floor stands
redefinition. A smart, interactive floor-equipped room can literally come to life and, in the process,
transform our life-world.
2.5. Feedback and Learning: Architecture as Immersive Interface
Feedback loops, realtime processing, and systemic learning are the basic characteristics of a smart
cybernetic system. It is believed that in less than a decade, more than one billion people all over the world
will be spending half as many hours in front of the computer screen as they will in physical space. de
Kerckhove, 2001). The computer screen has become the 17” gateway to all the digital information out
there. What if we are able to expand the ways by which we see, hear, touch and sense information? What
if we can release more people from the screen for more hours by distributing the interface around the
architectural environment? What if the walls, floors, lighting, ventilation and other facets of the
architectural environment begin to communicate information to the user? What if architecture as a whole
becomes a gigantic immersive interface to send and receive information? What if architecture becomes a
spatial synesthetic pump to channel, amplify and process the feedback loops of digital information?
Acoustic Tap Tracking system, developed by the team led by Joseph Paradiso of MIT Media Lab, is
capable of transforming any large flat surface such as a wall or a window or a table into a tap sensing
Page 9 of 13
interactive responsive surface. Paradiso, 2000). In distinction, the other higher tech devices developed
elsewhere that use optical cameras, touch screens, lasers, pens and light curtains, have many inherent
problems. From cost to portability, from effectiveness to scalability, these systems have not had much
success in transforming things at architectural scale.
Figures 11 and 12: Acoustic Tap Tracking System by Responsive Environments Group, MIT Media Lab
The ingenuity of Paradiso’s system is that it is buildable for under $500 and can be scaled to be installed on
large architectural surfaces. Earlier version of this device came from Paradiso’s collaborations with Media
Lab’s Tangible Media group, which resulted in the so-called PingPongPlus demonstrated at SIGGRAPH
1998. In the experimental prototype, four transducers sense the taps on a 4’x8’ shatterproof glass pane. The
analog audio signals are analyzed for peak timing and fed into the computer as intelligible taps such as
mouse clicks. Thus, it is possible to transform any shop front window into a simple interactive screen. If we
stretch our imagination around this invention, we can think of walls and tables becoming interactive
interfaces. The dumb, blank, abstract walls that divide can become interactive, intelligent, sensing, smart
surfaces that could truly connect in real-time. Information becomes immersive. Walls could become portals
to spatially distant lands or culturally other worlds.
Among other examples, The Tangible Media Group of MIT has demonstrated that more than simple bits of
information can be communicated through architectural interface. The group created the so-called
ambientROOM, which is remotely connected to a number of devices. Ishii, 1998). In the Active Wallpaper
project, sensors in a remote location pickup the activity level and transmit it to the ambientROOM where
the information is revealed in the form of a pattern of illuminated patches. When the activity is high, the
Page 10 of 13
movement of the patches is high and vice versa. In the Pinwheel project, colorful pinwheels spin at
different speeds based upon their information input. Ishii points out that these efforts “envision that the
architectural space we inhabit will be a new form of interface between humans and online digital
information”. Ishii, 1998). He sums up the project thus: “the ambientROOM surrounds the user within an
augmented environment – ‘putting the user inside the computer’ – by providing subtle, cognitively
background augmentations to activities conducted within the room.”
Teleporting a comprehensive experience is at the heart of these experiments. Two or more spaces, in this
scenario, could be in sync with each other despite their physical separation by vast distances, presumably
even interplanetary distances! This kind of connectivity radically redraws the map of spatially fragmented
geographic locations into a map of temporally contiguous experiences. Smart systems such as these,
question our outmoded notions about context, region, and neighborhood.
3. Epilogue
This paper serves as one of the first attempts to understand, define and frame smart architecture from a
complex-adaptive systems approach. The paper raises important questions about the meaning of smartness
in architecture and proposes the need for Smartness Quotient. SQ) and Building Operating System. BOS)
as a way to measure, standardize and systematize smart architecture. The paper also attempts to bring
together the camps of performative, kinetic, responsive and adaptive environments under the rubric of
smart architecture.
References
Born, M. 1962. Einstein’s Theory of Relativity. New York, NY: Dover Publications, Inc.
Cilliers, P. 1998. Complexity and Postmodernism. London: Routledge.
de Chardin, T. 1961. The Phenomenon of Man. New York, NY: Harper & Row.
de Kerckhove, D. 2001. The Architecture of Intelligence. Basel, Boston and Berlin: Birkhäuser.
Grosz, E. 2001. Architecture From the Outside. Cambridge, Mass: MIT Press.
Ibelings, H. ed. 2000. The Artificial Landscape. Rotterdam, Holland: NAI Publishers.
Page 11 of 13
Ishii, H, et al.. 1998. “Ambient Displays: Turning Architectural Space into an Interface between People and
Digital Information” in the Proceedings of the First International Workshop on Cooperative
Buildings. CoBuild '98., Springer Press, Darmstadt, Germany, pp. 22-32
Kelly, K. 1994. Out of Control: The New Biology of Machines. New York, NY: Perseus Books.
Mitchell, W. 1999. E-topia: Urban Life Jim, But Not As We Know It. Cambridge: The MIT Press.
Paradiso, J et al. . 2000. “Sensor systems for interactive surfaces” in IBM Systems Journal, Volume 39,
Numbers 3 & 4
Senagala, M. 2003. “Postspatial Architectures: The Emergence of Time-like Parametric Worlds” in Cultura
Digital y Diferenciacion, Proceedings of the SIGRADI International Conference, Rosario,
Argentina, pp. 375-382
Virilio, P. 1991. Lost Dimension. New York, NY: Semiotext(e).
Zellner, P. 1999. Hybrid Space. New York, NY: Rizzoli.
Kienzl, N. 2002. Evaluating Dynamic Building Materials: The Potential Impact of Climatically Responsive
Building Enclosures. Doctoral Thesis, Cambridge, Mass.: Harvard University
Pagani, F. 1999. Adaptive Buildings through Evolutionary Design. Doctoral Thesis, Vancouver, Canada:
The University of British Columbia
Park, C. 2003. Occupant Responsive Optimal Control of Smart Façade Systems. Doctoral Thesis, Atlanta,
Georgia Institute of Technology, USA
Potts, W. 1984. “The chorus line hypothesis of manoeuvre coordination in avian flocks" Nature, vol. 309,
pp. 344-345
Daniels, K. 1997. The Technology of Ecological Building. Verlag, Basel, Switzerland: Birkhauser.
Zuk, W. 1970. Kinetic Architecture. New York, NY: Van Nostrand Reinhold.
Author Information
Mahesh Senagala
Associate Professor and Associate Dean for Research
Page 12 of 13
College of Architecture
University of Texas at San Antonio, USA
[email protected]
Areas of Interest: Complex-adaptive systems, Tensile Fabric Structures, Cybernetics, Smart Architecture,
Critical Theory
Page 13 of 13