Urban Pixels: Painting the City with Light

Urban Pixels: Painting the City with Light Susanne Seitinger, Daniel S. Perry, William J. Mitchell

Smart Cities Group MIT Media Laboratory

E15, 20 Ames Street, Cambridge, MA 02142 susannes, dperry, wjm [@mit.edu]

ABSTRACT Urban environments are increasingly filled with digital display systems that are inflexible, flat, bounded, high-resolution, and unresponsive. In this paper, we explore the potential of physically instantiated pixels that enable flexible, reconfigurable, unbounded, low-resolution, and responsive urban displays. Urban Pixels are nodes in a wireless network of physical pixels for urban spaces. Each pixel unit includes a microcontroller, RF transceiver (433 MHz), LED module (ten bright, white LEDs), rechargeable Li-Ion battery pack, IR sensor and renewable energy source such as photo-voltaic cells. Two acrylic half-spheres (4-inch diameter) protect the components from the elements. No additional wiring is needed for communication and the units can be mounted individually to any surface. A small-scale prototype network of fifty Urban Pixels was displayed on a façade of Eden Court Theater in Inverness, Scotland from June 1 – June 7, 2008. The public was encouraged to change display patterns via SMS or to interact with individual units via flashlights. We observed and informally interviewed theater guests and passers-by interacting with the façade for several nights. Based on these results, we outline an exciting problem space for designing displays and lighting systems in cities.

Author Keywords urban display, lighting, ambient media, ubiquitous computing, urban computing, interaction design

ACM Classification Keywords H5.m. Information interfaces and presentation (e.g., HCI): Miscellaneous.

INTRODUCTION Even though urban displays have become increasingly sophisticated, the predominant designs are inflexible, flat and bounded. These screens support high-resolution content that often requires a user’s undivided attention and would be more appropriate for an indoor, private setting. While

urban lighting systems have become increasingly sophisticated, these systems do not enable communication or interactivity. There are also commercial examples of environments with sophisticated public display and lighting systems such as Freemont Street in Las Vegas or parts of Times Square in New York and more recently a shopping district in Beijing, China where the world’s largest ceiling display was unveiled. These custom installations are not ubiquitous and often rely on complex, centralized control systems.

We address the gaps in existing systems described above by blurring the boundaries between urban display and lighting systems. Urban Pixels are networks of physical pixels that

Permission to make digital or hard copies of all or part of this work forpersonal or classroom use is granted without fee provided that copies arenot made or distributed for profit or commercial advantage and that copiesbear this notice and the full citation on the first page. To copy otherwise,or republish, to post on servers or to redistribute to lists, requires priorspecific permission and/or a fee. CHI 2009, April 4–9, 2009, Boston, Massachusetts, USA. Copyright 2009 ACM 978-1-60558-246-7/09/04...$5.00.

Figure 1. Urban Pixels installed at Eden Court Theater, Inverness, Scotland, June 2008.

CHI 2009 ~ Art Creation April 7th, 2009 ~ Boston, MA, USA

839

can be placed flexibly in any configuration on any surface in the city. They support low-resolution, context-sensitive ambient displays. Users may interact with the units by placing them spontaneously, by triggering sensors or by sending messages from their mobile devices. Together, these capabilities present design criteria for a new generation of urban display and lighting technologies.

We developed a small-scale prototype network of fifty Urban Pixels to demonstrate the importance of exploring new approaches to urban displays and lighting (see Figure 1). After providing some inspiration for this work from diverse fields, we describe the technical implementation, related HCI research, the impact of Urban Pixels. We conclude by outlining trajectories for designing urban displays, ambient information systems and interactive, responsive lighting in the future.

PAINTING THE CITY WITH LIGHT In order to develop new strategies for deploying urban displays, we draw on diverse examples from art and photography and previous taxonomies for ambient display technologies [44]. First, we explore the lessons from painting, photography and cinematography. Second, we review a series of inspiring examples.

At the urban scale, strategically deploying ambient light makes the night-time city landscape editable. When a background is completely dark, only highlighted features remain visible while other characteristics blend into the background. Pioneering photographers such as Stieglitz identified the power of this strategy: “Such imperfections (like halations) introduced (…) life into nighttime images and recreated what the photographer saw as he exposed the image. This was ‘real picture-making,’ as opposed to a mere topographical view.” [34, p.69] For Stieglitz, the flexibility of the night-time setting unleashed the photographer’s true creativity. For example, these techniques enable photographers and filmmakers to recast spatial cues in static and moving images. In Film Sense, Eisenstein describes the effect, which he perceptively links to electrification: “All sense of perspective and of realistic depth is washed away by a nocturnal sea of electric advertising. Far and near, small (in the foreground) and large (in the background), soaring aloft and dying away (…) these lights tend to abolish all sense of real space, finally melting into a single plane of colored light points and neon lines moving over a surface of black velvet sky.” [13, p.98] For daytime settings, painters such as William Turner and later the Impressionists also recognized the importance of light for an outdoor scene. Applying dynamic digital technologies in the spirit of these painters, photographers and cinematographers, designers and technologists can begin unlock the full potential for deploying points of programmable light in the city.

Figure 2. Blinkenlights installation at Toronto City Hall, Canada, 2008. Excerpt of photograph by

http://www.flickr.com/photos/wvs/2909334414/ reproduced under Creative Commons License, http://creativecommons.org/licenses/by-nc/2.0/

Figure 3. Laser Tag on a building façade. Photograph by http://www.flickr.com/photos/urban_data/396087351/

reproduced under Creative Commons License, http://creativecommons.org/licenses/by/2.0/

Figure 4. Tram with Throwies. Excerpt of photograph by http://flickr.com/photos/urban_data/243453680/

reproduced under Creative Commons License http://creativecommons.org/licenses/by/2.0/

CHI 2009 ~ Art Creation April 7th, 2009 ~ Boston, MA, USA

840

Several examples from art and architecture demonstrate how strategically placing light in the city at night expands the design opportunities for urban displays. While these examples are not intended as an exhaustive list, they provide a provocative set of strategies for distributing programmable points of light in various contexts.

Using the building as interface: Blinkenlights transformed the façade of the Haus des Lehrers in Berlin into the world’s biggest interactive computer display. People could also create their own animations such as the heart shown here via a program called Blinkenpaint. [5] The following figure shows a similar installation on the Toronto City Hall. (see Figure 2)

Appropriating public facades through projection: Laser Tag by Graffiti Research Lab allows people to project graffiti onto a building. [17] (see Figure 3)

Interacting with façades through projection: Body Movies by Lozano-Hemmer are interactive projections that composite shadows of people currently in the plaza with portraits taken on the streets of the city. [28] (see Figure 5)

Deploying individual pixels: Throwies by Graffiti Research Lab are deployable pixels consisting of a coin-cell battery, 10mm diffused LED and magnet. [18] (see Figure 4)

Connecting individual pixels: Supercluster by Leo Villareal is an early example of a building-scale digitally programmable façade. The system consists of 640 LEDs distributed across a 45 by 120 foot display that covered the entire façade of PS 1 museum in New York. [47] (see Figure 6)

Transforming architecture into ambient display: At the Allianz Arena, Herzog and de Meuron created façade of inflated ETFE-foil air panels that can be transformed with colored light. The ambient display usually signals the colors of the competing teams. [20] (see Figure 7)

Creating landscapes with pixels: White Noise White Light, Meejin Yoon was shown at the Athens Olympics 2004. The piece demonstrates the transformative effect of light in a landscape and as a new landscape to be inhabited by visitors. [51] (see Figure 8) These examples demonstrate how to translate the painterly strategies described above into display and lighting systems. They transform the city into a canvas upon which to distribute points of light. Traditional and commercial systems do not mobilize many of the strategies used in the examples above such as flexible placement, low resolution or interactivity.

From these examples, we distill the following key features as design criteria for Urban Pixels:

• flexible placement • autonomous power

Figure 7. Allianzarena in blue, Munich, Germany. Photograph by

http://www.flickr.com/photos/hcii/61614000/ reproduced under Creative Commons License,

http://creativecommons.org/licenses/by-nc-nd/2.0/deed.en

Figure 6. Leo Villareal, Supercluster, 2003. White LEDs, custom software, electrical hardware, suspension material. Installation view. Image

courtesy of Gering & Lopez Gallery, New York.

Figure 5. Rafael Lozano-Hemmer, "Body Movies" (2002). Linz, Austria. Photo by

Antimodular Research.

CHI 2009 ~ Art Creation April 7th, 2009 ~ Boston, MA, USA

841

• unbounded • variable resolution • responsive A system should support flexible placement of the individual points of light such as Throwies or White Noise White Light. Such flexibility implies a source of on-board power. The network should support an unbounded display that is not pre-framed or flat just like the projects which employ projection onto building façades. In other words, a system should adapt to its substrate, for example a building façade, a landscape or any other surface. Flexibility makes it possible to vary the display resolution throughout the network and communicate with ambient messages. Spontaneous placement allows users to interact directly with individual pixels. Additional sensors and communications enable further responsive behaviors. We believe that these criteria present a more inclusive approach to urban lighting and display technology that will enable designers and users to paint the city with light.

TECHNICAL IMPLEMENTATION

Urban Pixel Units Urban Pixels presents a wireless network of physical pixels for urban spaces. (see Figure 9) Each unit includes a CC1010 microcontroller and RF transceiver (433 MHz), LED module (ten bright, white LEDs), 2.5mm power plug for charging, rechargeable 3.7V, 66000 mAh Li-Ion battery pack, IR sensor and renewable energy source such as photo-voltaic cells. The communications hardware and firmware is modeled on the RFRAIN nodes [24]. We have expanded each node to include an LED module, IR sensor and larger battery pack for longer deployment and brighter display capabilities. Two 4-inch diameter acrylic half-spheres protect the components from the elements. They are held in place by a circular rib structure water-jet cut in polycarbonate. One half-sphere has been sandblasted and lined with a layer of 1/8” stuffing material to diffuse the light emanating from the individual LEDs. The units can be opened easily by hand with a thumb screw. Each unit can

be connected via a 6/32” threaded screw to various connectors depending on the surface type including brackets, suction cups, and others. No additional wiring is needed for communication so that the units can be mounted individually to any surface (see Figure 9).

Communication Protocol The current single-hop network with centralized control facilitates pattern generation for flexible arrangements of units. Unlike most distributed sensor networks, each node in Urban Pixels is unaware of the other nodes. When initially programmed, nodes in the network receive a unique address in order to uniquely identify it, but do not receive information about the number of other nodes or structure of the network. A centralized node broadcasts all necessary information about the network to all of the other nodes. A laptop controls the entire network via a standard RS-232 serial connection with the centralized node. Thus, the structure of the lighting network can be changed with a laptop without necessitating the reprogramming of all of the distributed nodes. Since the number of patterns the network is capable of displaying is substantially less then the number of nodes in the network, it is more efficient to encode patterns. The encoding of messages depends on the pattern type. Units receive a pattern type (see “Interactions” below) and parse unique address, each node is not individually addressable in the network.

IR Sensors It is critical that the lights for a distributed lighting network do not affect the network's light sensing capabilities. A node in a lighting network that operates only at night should not shut off because other nodes in the network are lit. The IR sensor detects radiation in the 850 nm range which is on the fringe of the visible light spectrum. Because of this, the sensor can detect any light source containing a substantial amount of red light (the sun, incandescent light bulbs, etc.).

Since the bright, white LEDs output light in the upper range of the visible light spectrum, approximately 500 nm, they do not effect the on board IR sensors. The IR sensor was connected to one of the several 10-bit ADC channels on the CC1010 without any amplifying circuitry. Initial tests indicated that the IR sensors on each node could detect an incandescent flashlight at a range of 20 ft. The Urban Pixels needed to have the ability to dynamically adjust to different environments without being reprogrammed which necessitated self-calibrating nodes. At initialization each node obtains several samples from the IR sensor. These samples are stored and averaged in order to obtain a threshold, which can be different for each node. In addition to this, each node periodically re-calibrates and updates its threshold in order to dynamically adjust to the environment.

Interactions In the current system, people can interact with Urban Pixels in three ways. First, they can control the entire display from a base station connected to a laptop computer. The interface

Figure 8. White Noise White Light. Photo courtesy of courtesy of J. Meejin Yoon / MY

Studio.

CHI 2009 ~ Art Creation April 7th, 2009 ~ Boston, MA, USA

842

allows the user to reconfigure the number of rows and columns, change the display frequency, reset the units and change patterns. The available patterns were direct interaction mode, random flashing, vertical lines, horizontal lines, and rain.

Second, they can send an SMS single-digit code to a GSM modem that changes the pattern. The SMS system was handled by an adaptation of a commercial product developed by Richard Wilson of Distance Lab, Forres, Scotland. The same patterns were available as displayed in the interface above. Finally, people can control individual units with a flashlight via the IR sensor when they are set to direct interaction mode (see Figures 10, 11).

RELATED WORK In addition to the media installations and façades described above, the design and implementation of Urban Pixels draws on three bodies of related research in tangible computing, wireless networks and ambient information display.

Physical Pixels The building blocks for an Urban Pixels network are physical pixels, tangible instantiations the pixels on a computer screen [15,21,22]. The following list provides a brief history of the evolution of physical pixels:

• I/O Bulb and Luminous Room [46] • Nami Project [19] • DataTiles [39] • Pushpin Computing [7] • SoundMites [4] • Siftables [30] • Paintable Computing [7] Some of these examples such as DataTiles [39] and Siftables [30] include high-resolution displays that convey meaning individually and as a network. Other examples such as Pushpins [7] and Namis [19] represent individual pixels that convey meaning as a network through low-resolution, ambient visual displays. Unlike its predecessors, Urban Pixels are specifically intended for urban scale applications that span a building or several buildings. Their relationship with the urban context unlocks the potential for exploiting all three dimensions of space while most preceding work is limited to table-top applications.

Wireless Sensor Networks The notion of sensor networks is approximately 10 yrs old. [14] The first reference to sensor networks was 30 years ago in “Distributed Sensor networks” Carnegie Mellon University Workshop proceedings. Dec. 1978 [26]. According to Akyildiz [1], the basic requirements for a sensor network are: • support very large numbers of unattended nodes • adaptive to environment • adjust to unpredictable task dynamics The envisioned applications for sensor networks are often in hostile environments such as war zones or post-emergency situations, but there are projects in the smart home/smart environment area and for urban environments [33,38,40] as well. The application spaces identified for wireless sensor networks include: passive habitat monitoring, asset tracking, smart environments and emergency zone monitoring. [1,14,41,50] Networks of physical pixels do not necessarily include sensors, however, they are smart devices that are also aware of their surroundings and can thus be thought of as sensor networks. In addition, they have the added feature of an

Figure 9. Urban Pixel prototypes.

Figure 11. Painting with light by using a flashlight to turn individual Urban Pixels on and off.

Figure 10. Urban Pixels interface for changing patterns.

CHI 2009 ~ Art Creation April 7th, 2009 ~ Boston, MA, USA

843

output function through a low-resolution LED blinking or a high-resolution screen such as in the Siftables. [30] In his work on the Plug sensor network Lifton found that browsing sensor network data is a considerable challenge and we can imagine that on-board feedback devices could assist in those instances where sensor network nodes are in sight. [26,27]

The typical parameters for designing sensor network systems include: fault tolerance, scalability, production costs, hardware constraints, network topology, environment, transmission medium, and power consumption. [1,2,14,41,50] While these parameters are equally important in Urban Pixels, the current system emphasizes central control over distributed, redundant controls. In this sense, Urban Pixels remains closer to typical display technologies than distributed systems.

Ambient Information Systems Ambient information systems exploit the power of our split attention between center and periphery. [2,35,48] In his seminal articles on ubiquitous computing, Weiser described the Dangling String installation by artist Natalie Jeremijenko to illustrate the power of ambient displays. The movement of the string gave people an instant sense of network traffic in the building without requiring significant attention on their part. [49]

Tangible computing [15,21,22] also includes the premise that tangible objects are coupled with calm, peripheral displays that do not interfere with the central object of inquiry. Ishii and Ullmer's showed some examples using projection and other media in AmbientROOM [46]. While there have been proposals for ambient media in urban settings [9,16], there is a need for more exploration in this domain. Urban Pixels provide an infrastructure for ambient information systems in any setting.

URBAN PIXELS IN THE WILD A network of 50 Urban Pixels was displayed on a façade of Eden Court Theater the premiere cultural venue in Inverness, Scotland from June 1-June 7, 2008. The system was running every evening from approximately 10pm until midnight when the theater closed. Fliers were distributed throughout the building explaining the SMS codes for the available patterns.

There was a public opening on June 3, 2008 that was attended by approximately 50 people. During the opening, the research team assisted guests with the SMS system or flashlights. Throughout the week, the public was encouraged to change display patterns via SMS or to interact with individual units via flashlights. We observed and informally interviewed guests and passers-by interacting with the façade.

DISCUSSION Looking back on the design criteria enumerated above (flexible placement, autonomous power, unbounded,

D

variable resolution, responsive) Urban Pixels fared well during its week-long deployment.

Flexible Placement, Unbounded Frame and Variable Resolution – A Liberated Infrastructure for Urban Display and Lighting First, Urban Pixels was easy to deploy requiring less than one hour and two people to mount on the building façade. The metal brackets and zip-ties also allowed for easy reconfigurability of the display arrangement (see Figure 12).

Second, some units were distributed throughout the theater grounds where people could touch them and move them around. This flexibility links Urban Pixels to traditions of lantern festivals all around the world with the additional characteristic of programmability.

Third, the low resolution nature of the display proved surprising for many people: “Why was there no text? Why couldn’t they write a message for their friends?” After recovering from their initial surprise, onlookers enjoyed the calm nature of the installation.

Fourth, Urban Pixels enlivened the façade of Eden Court throughout its week-long deployment without detracting from the building. The ability to layer ambient information onto an existing structure proved very powerful. People still recognized their theater shining through the light. The round form factor and mounting strategy further enhanced the hovering appearance of the system in front of the architecture.

Figure 12. Urban Pixels mounting system.

CHI 2009 ~ Art Creation April 7th, 2009 ~ Boston, MA, USA

844

Painting with Light The late sunsets and long dusk in Scotland in June provided particularly interesting conditions for experimenting with the mixing of digital and natural lighting. The Pixel units were bright enough to be visible during early evening and then really became visible late at night when the building faded into an invisible night-time background (see Figure 13). These transitions were particularly interesting because they showed how the system was constantly in conversation with the physical environment.

Painters such as William Turner (see Figure 15) and the Impressionists tried to capture these effects on the canvas through the strategic placement of light and dark zones. Digital programmability confounds traditional mechanisms for matching message and substrate because of the replicability of digital data [31]. Zooming into a digital image does not reveal additional details beyond the point where the grainy pixilated structure emerges. A painting cannot be reduced to its component parts in the same way.

Urban Pixels recaptures these painterly techniques three-dimensional display networks of digitally programmable points of light. As mentioned above, the notion of physical pixels can be traced back to ubiquitous computing [8], tangible computing [15,21,22,46] and ambient displays [44]. By exploiting humans’ peripheral perception, physical ambient display systems expand our ability to glean information from the environment. Urban Pixels couples the power of physical pixels with the painterly strategies for capturing a mood or transforming a landscape-scene through light.

Responsive – Interacting with Urban Pixels Direct interaction using flashlights to turn individual pixels on proved to be very enjoyable for people (see Figures 11, 14). The clear connection between cause and effect seemed to facilitate a more personal and social set of interactions around the lighting system and façade [43]. Experiences with other installations such as the Tactile Luminous Floor [10] have shown the potential for enhancing people’s experience of space through an ambient lighting system.

The SMS system gave full control to users, which was very popular. However, they could never be entirely certain when their SMS had reached the system. Transitions from pattern to pattern happened quickly and there were many repeat requests that led to longer stretches of the same pattern rather than more rapid transitions among patterns.

Many people also simply enjoyed the display from afar in its default mode cycling through all of the possible display patterns and frequencies. The fast-paced transitions of patterns seemed to elicit more enjoyment similar to the effect of firework displays.

These early interaction examples only hint at the many possible responses that Urban Pixels could elicit among people. There are many more social interactions that could result from placing Urban Pixels in different urban contexts

Additional systematic user studies should be made to expand the validity of Urban Pixels as an approach to innovative, low-resolution urban interactive displays.

Flexible Placement and Autonomous Power – Future Technical Requirements The technical implementation functioned reliably throughout the week-long test period. Future systems will provide increased redundancy, localization, multi-hop communication, and the ability to individually address all nodes. The implementation confirmed the ease of use and

Figure 13. As night falls, pixels become more prominent than architecture.

Figure 14. Passersby viewing Urban Pixels. Visible traces (bicycles) of everyday urban life.

Figure 15. J.M.W. Turner, "Yacht Approaching the Coast" (ca. 1840-1845). Oil on canvas. Photograph ©

Tate, London 2008.

CHI 2009 ~ Art Creation April 7th, 2009 ~ Boston, MA, USA

845

the benefits of applying wireless sensor network techniques to a display system. [11,41]

The expressivity of the system would be significantly enhanced by full-color lighting modules. The possibility of interacting with natural light conditions would be greatly enhanced. Also full-color would enable a broader range of signaling despite the low resolution.

Wireless communication and battery packs supported free and autonomous placement for the duration of the installation. For long-term deployments or inaccessible locations, an on-board renewable energy source would support even more flexible placement of pixels. Previous design iterations have incorporated PV cells into the pixel units (see Figure 16).

Other solutions such as wind generators and wave or river currents are being studied. Power challenges are typical of any wireless network and may require different combinations of power and lighting modules to accommodate diverse application scenarios. There have been scientific [41] and artistic [6] experiments with power harvesting. And these questions will continue to play a key role in advancing ubiquitous computing infrastructures [50].

CONCLUSION Urban Pixels open a new problem space for urban display and information systems that are flexible, unbounded, and low-resolution. The temporary light installation was spontaneously placed on a theater façade without support infrastructure. The network of pixels enhanced the building leaving no traces behind after its deployment. The interactions between changing natural conditions and the lighting units enriched the preprogrammed display patterns significantly. Together, the natural and programmed patterns demonstrated the merits of a painterly approach to deploying points of light in an urban scene that could be explored further. Visitors to the theater enjoyed interacting with the system and the visual effect it had on the theater.

Based on our explorations, we believe that systems like Urban Pixels can enable new interactions among urbanites, information, displays, and light as proposed by many visions for new kinds of urban spaces [29]. They can support user-deployment as well as private- or public-sector deployments. People will be able to reprogram these systems directly, via sensor interactions or by simply

experiencing them in a dynamic urban setting. More experimentation is required to explore the full potential of these “liberated” infrastructures that blur the boundary between urban displays, ambient information systems and traditional infrastructures such as urban street-lighting. [24] The social potential [43] of these infrastructures also promises to be a rich domain for future work. [22,31,42]

ACKNOWLEDGMENTS Thanks to the anonymous reviewers for their constructive and encouraging critiques. Photographs of the installation are by the authors, Matthew Karau, and Richard Wilson. Thanks to everyone who assisted with the development of Urban Pixels especially Peter Schmitt, Mark Feldmeier, Ellen Yi Chen, and Franco Vairani. Special thanks are due to Richard Wilson (Distance Lab, Forres, Scotland) for interfacing Urban Pixels with his SMS system. Thanks to our many other supporters in Scotland: MIT Media Laboratory sponsor Highlands and Islands Enterprise in particular Douglas Yule, Laura Dingwall, and Stephanie Anderson as well as Rab Gordon (Rainnea Ltd.) and the whole team at Eden Court Theater. We are grateful to our MT Media Laboratory Sponsor Consortia for supporting this work.

REFERENCES 1. Akyildiz, I.F., Su, W., Sankarasubramaniam, Y.,

Cayirci, E. A survey on sensor networks. IEEE Communications Magazine 40, 8 (August 2002), 102-114.

2. Ambient Intelligence. Philips Research Technology Magazine. Password 23 (May 2005).

3. Bellotti, V. et al. Making sense of sensing systems: Five questions for designers and researchers. In Proc. CHI 2002, ACM Press (2002), pp-pp.

4. Bouchard, D. and Maes, P. SoundMites: Embodied Emergence. http://ambient.media.mit.edu/projects.php?action=details&id=41

5. BlinkenLights. Berlin, 2001-2002. www.blinkenlights.de.

6. Bruges, J. Wind to Light. London, 2007. http://www.jasonbruges.com/

7. Butera, W. Text display and graphics control on a paintable computer. In Proc. of the int’l Conference on Self-Adaptive and Self-Organizing Systems (2007), 45-54.

8. Buxton, W. Absorbing and squeezing out: On sponges and ubiquitous computing. In Proc. of the International Broadcasting Symposium 16 (November 13-16), 91-96.

9. Chatzitsakyris, P., Ducla-Soares, G., and Zulas, A. IN-Visible: perceiving invisible urban information through ambient media. Ext. Abstracts CHI 2004, ACM Press (2004), 1348-1350.

Figure 16. Previous design iterations for Urban Pixels that incorporate PV cells.

CHI 2009 ~ Art Creation April 7th, 2009 ~ Boston, MA, USA

846

10. Delbrück, T., Whatley, A. M., Douglas, R., Eng, K., Hepp, K., and Verschure, P. F. A tactile luminous floor for an interactive autonomous space. Robotics and Autonomous Systems 55, 6 (2007), 433-443.

11. Dey, A., Mankoff, J., Abowd, G., and Carter, S. Distributed mediation of ambiguous context in aware environments. In Proc. of the ACM Symposium on User Interface Software and Technology (2002), 121-130.

12. Ducatel, K., Bogdanowicz, M., Scapolo, F., Leijten, J. and Burgelman, J.-C. Scenarios for ambient intelligence in 2010. Final Report European Commission IST Advisory Group (2001).

13. Eisenstein, S. M. The Film Sense. Jay Leyda, transl. and ed. Harvest Book, Harcourt, Brace & World, New York, 1942, 1975.

14. Estrin, D., Govindan, R., Heidemann, J. S., and Kumar S. Next century challenges: Scalable coordination in sensor networks. In Mobile Computing and Networking, ACM Press (1999), 263–270.

15. Fitzmaurice, G., Ishii, H., and Buxton, W. Bricks: Laying the foundations for graspable user interfaces. In Proc. of CHI 1995, ACM Press (1995), 442-449.

16. Frenchman, D., Rojas, F. Zaragoza’s Digital Mile: Place-Making in a new public realm. Places 18, 2 (2005), 16-25.

17. Graffiti Research Lab. Laser Tag. 2007. http://graffitiresearchlab.com/?page_id=76#video

18. Graffiti Research Lab. Throwies. http://graffitiresearchlab.com/?page_id=6

19. Heaton, K. B., Poor, R. D., and Wheeler, A. J. Nami. In Conference Abstracts and Applications SIGGRAPH 1999, ACM Press (1999), 215.

20. Herzog, J. and De Meuron, P. Allianz Arena. Munich, 2005. http://en.wikipedia.org/wiki/Allianz_Arena

21. Ishii, H. Tangible pixels: Beyond pixels. In Proc. of TEI 2008, ACM Press (2008), XV-XXV.

22. Ishii, H. and Ullmer, B. Tangible bits: Towards seamless interfaces between people, bits and atoms. In Proc. of CHI 1997, ACM Press (1997), 234-241.

23. Ito, M. D. and Okabe, M. eds. Personal, Portable, Pedestrian: Mobile Phones in Japanese Life. MIT Press, Cambridge, MA, 2005.

24. Jakle, J.A. City Lights: Illuminating the American Night. Johns Hopkins University Press, Baltimore, London, 2001.

25. Laibowitz, M. and Paradiso, J.A. Wireless wearable transceivers. Circuit Cellar 163 (2004), 28-39.

26. Lifton, J., Feldmeier, M., Ono, Y., Lewis, C., and Paradiso, J.A. A platform for ubiquitous sensor deployment in occupational and domestic environments. In Proc. of IPSN, ACM Press (2007), 119-127.

27. Lifton, J., Broxton, M., and Paradiso, J.A. Experiences and directions in pushpin computing. In Proc. of IPSN, ACM Press (2005), 416-421

28. Lozano Hemmer, R. Body Movies. Rotterdam, 2001. www.lozano-hemmer.com/video/bodymovies.html

29. McCullough, M. Digital Ground: Architecture, Pervasive Computing, and Environmental Knowing. MIT Press, Cambridge, MA, 2004.

30. Merrill, D., Kalanithi, J., and Maes, P. Siftables: towards sensor network user interfaces. In Proc. of TEI 2007, ACM Press (2007), 75-78.

31. Mitchell, W. J. The Reconfigured Eye. MIT Press, Cambridge, MA, 1992.

32. Murthy, R. Street Media: Ambient Messages in an Urban Space. MIT SM Thesis, 2005.

33. Murty, R., Gosain, A., Tierney, M., Brody, A., Fahad, A., Bers, J. and Welsh, M. Harvard University Technical Report, TR-13-07 (September, 2007). http://www.citysense.net/

34. Neumann, D. Architecture of the Night: The Illuminated Building. Presetel, Munich, New York, 2002.

35. Norman, D. Invisible Computer. MIT Press, Cambridge, MA, 1998.

36. Nye, D. E. Electrifying America: Social Meanings of a New Technology. MIT Press, Cambridge, MA, 1992.

37. Paulos, E. “Urban Probes.” INTEL. http://urban-atmospheres.net/UrbanProbes

38. Pfeiffer, E. WhereWare. Technology Review (2003). 39. Rekimoto, J., Ullmer, B., and Oba, H. DataTiles: A

modular platform for mixed physical and graphical interactions. In Proceedings of the SIGCHI 2001 Conference on Human Factors in Computing Systems, ACM Press (2001), 269-276.

40. Riva, O. and Borcea, C. The Urbanet revolution: Sensor power to the people! IEEE Pervasive Computing (2007), 41-49.

41. Savvides, A., Girod, , L., Srivastava, M. and Estrin, D. Localization in sensor networks. In Znati, Radhavendra and Sivalingam, eds. Wireless Sensor Networks. Springer, New York, 2006.

42. Schivelbusch, W. Licht, Schein und Wahn. ERCO Edition, Berlin, 1992.

43. Struppek, M. The social potential of urban screens. Visual Communication 5, 2 (2006): 173-188.

44. Tomitsch, M., Kappel, K., Lehner, A. and Grechenig, T. Towards a taxonomy for ambient information systems. In Pervasive 2007 W9 Ambient Information Systems (May 13, 2007).

45. Tscheligi, M. ed. Ambient intelligence: Exploring our living environment. Interactions 12, 4 (2005): Special Issue.

CHI 2009 ~ Art Creation April 7th, 2009 ~ Boston, MA, USA

847

46. Underkoffler, J., Ullmer, B. and Ishii, H. Emancipated pixels: Real-world graphics in the luminous room. In Proc. of SIGGRAPH 1999, ACM Press (1999), 385-392.

47. Villareal, L. Supercluster. New York, 2003-2004. http://www.villareal.net/project_3.html

48. Want, R. et al. Disappearing hardware. IEEE Pervasive Computing (Jan.-March 2002): 36-47.

49. Weiser, M. The computer for the 21st century. Scientific American (1991), 94-104.

50. Yick, J., Mukherjee, B., and Ghosal, D. Wireless sensor network survey. Comput. Netw. 52, 12 (Aug. 2008), 2292-2330.

51. Yoon, M. White Noise White Light. Athens, 2004. http://www.mystudio.us/

CHI 2009 ~ Art Creation April 7th, 2009 ~ Boston, MA, USA

848