1Introduction

As exemplified by Tomitsch and Verhoeff previously, the idea of augmenting public spaces with digital media, commonly referred to as Media Architecture, is no longer a future vision. We are now entering a time where actuators in the form of computer-controlled lighting and display elements are being deployed at large scales in our cities. At the same time, the coverage of high-speed internet allows manifold interaction opportunities between open data and citizens, thereby enabling the Internet of things in the public domain. In a previous chapter of this book, Tomitsch argued to shift away from the traditional view of media architecture taking the form of large-scale media façades. He introduced the notion of ‘city-apps’ that can take the form of any digital interface that provides real-time data or enables people to interact with the city’s infrastructure. With ‘DIY / DIWO Media Architecture’, Caldwell and Foth explored a bottom-up approach that is more open and accessible for laypeople with the purpose of fostering community engagement. Complementing the aforementioned research, in this chapter we focus on the question how situated data should be integrated and visualized in the urban environment. We share our practical experiences from exposing a display prototype to public places. The development of the prototype builds on a recently discussed do-it-yourself (DIY) toolkit for hybrid media surfaces combining low-resolution and high-resolution display techniques. Making our approach available to others, we aim to reduce the technical burdens that currently hinder the wider dissemination of alternative solutions to current city developments.

Urban interfaces in the form of instrumented buildings and public displays can provide additional communication channels that are not addressed to individuals but rather to the public with in-situ access. In current implementations, the mainstream approach is retrofitting the urban environment with giant-sized television (TV)-like screens to display ubiquitous advertising and to serve as large-scale billboards where content and information can be exchanged at a fast pace. Apart from a questionable aesthetic appearance and integration into the environment, these high-resolution screens demand constant foreground visual attention, as multiple information streams compete for attention, and they fail to take advantage of humans’ background processing capabilities. Ambient information as an overarching concept for more integrated, non-intrusive and responsive computing can be one alternative way of using media architecture by learning about people and their identities: their habits, preferences, behavioral patterns, etc., and how to apply such knowledge in varying contexts (Ishii et al., 1998). Following this design philosophy, however, poses several challenges; one includes the task of aesthetically integrating media architecture into physical surroundings and structures, which demands careful considerations of content, context, and users to create meaningful experiences within the built environment (Dalsgaard and Halskov, 2010). We consider it advantageous that media architecture is not just perceived as an information layer-on-top filled with advertising. Instead, in our everyday lives, media architecture can provide mutual benefits: Citizens can, for example, gain more awareness of their surroundings and become more informed about what is happening around them, which, in turn, can provoke the better use of common resources. Examples in this field range from citizens’ reported infrastructure issues (Korsgaard and Brynskov, 2014), empowering sharing economies among neighborhoods (Pop, 2014), or real-time data visualization in urban contexts (Calabrese and Ratti, 2006). However, one dominant question persists: In which way should such digital information be integrated and visualized in the built environment? In order to prevent increased distraction in cities due to information overload, the need exists to rethink information visualization in conjunction with media architectural displays, as “clutter and confusion are failures of design, not attributes of information.” (Tufte, 1992) In this vein, ambient information provides the benefit of a more seamless integration into the environment than do large-scale TV-like screens. Moreover, aesthetic animations can add further value, activating citizens’ peripheral attention. One way of making ambient information comprehensible to people is to utilize strong metaphors (Ishii et al., 1998). However, their range of options is unimaginably large. Therefore, former investigations dealt with finding the right target dimensions for ambient displays in the public domain (Wiethoff and Gehring, 2012).

In this chapter we share our experiences investigating the integration of media architecture. Our chapter includes our experiences exposing an ambient display to different locations (i.e., indoor vs. outdoor) and contexts (i.e., work vs. leisure). In doing so, we aimed to understand people’s perceptions and reactions of such a display type in the public domain.

1.1Integration of Media Architecture

For media architecture to be successfully integrated into physical surroundings and structures, it has to function on different levels (Wiethoff, 2013). Based on our previous research from conducting various media architectural interventions in the past years and building on research findings in the emerging media architecture community, we identified the following key aspects: (1) Media architecture has, on its most basic level, to deliver aesthetic architectural experiences in order to gain widespread acceptance (Hoggenmueller and Wiethoff, 2014) even if users are not in favor of or interested in the displayed information. The ambient nature of the content can be perceived as an added aesthetic value (Hoggenmueller and Wiethoff, 2015). (2) The encoding of digital information must not be of a high cognitive load (McCullough, 2013) so as to support the easy perception of digital information through the built environment (Dalsgaard and Halskov, 2010). (3) If content can be manipulated via interactivity, the interface should be barrier free and must not require additional alternative means to enable control. This matter is especially demanding in order to include all users and age groups present in the public domain (Wiethoff et al., 2014).

1.3Ambient information

“A persistent layer of messages for somebody else …” (McCullough, 2013)

Ambient information systems take advantage of humans’ background processing capabilities. Ishii et al. first described the vision of an architectural space that serves as an interface for displaying information in the periphery of the user’s attention (Ishii et al., 1998). Inspired by natural phenomena, such as wind, sunlight, or temperature, ambient media displays were embedded into the physical environment using subtle changes to process information. Using similar metaphors, the Power-Aware-Cord, a re-design of a common-electrical power strip, encodes the current energy consumption using light intensity as a visual variable (Gustafsson and Gyllenswärd, 2005). In general, ambient displays feature high aesthetic qualities and support the monitoring of non-critical information. However, some systems also provide alerts to inform the user about important changes. Pousman et al. (2006) described the degree at which an ambient information system forces users to interrupt their primary tasks on a notification level.

Applying “ambient media” to urban contexts has been recently discussed by various researchers (Chatzitsakyris et al., 2004; Colangelo, 2014; Sade, 2014; Seitinger et al., 2009). They stress that predominant high-resolution public displays are competing for the user’s undivided attention and cannot be integrated in an architecturally friendly way. Seitinger et al. (2009) therefore proposed a network of wireless, solar-powered lighting units, called Urban pixels, which can be flexibly arranged on any surface (see Figure 2). These new approaches explore the transition among urban displays, ambient information systems, and street lighting. They are also inspired by already-realized cases of media architecture. One prominent example is the Allianz Arena in Munich (see Figure 1), of which the outer shell can be transformed into an ambient display to monitor what is going on inside: using color and patterns as ambient media information. One limitation coming with the low-resolution nature of ambient media is the encoding of information in a way that large audiences can understand: While locals and soccer fans can decode the meaning here, it might be more difficult for visitors to understand the information. In this vein, Offenhuber and Seitinger (2014) stress the arbitrariness of visual encoding when communicating information via low-resolution media architecture. Claiming that “content follows resolution,” they discuss five visual variables, including color, movement, text, images, and the intrinsic shape of the building, to encode information depending on the infrastructure. In order to convey explicit information via low-resolution displays, they investigate the use of a mobile interface as an additional layer for annotations, which does not, however, enable barrier-free use. Another example of ambient information visualization in the urban realm is a conceptualized art installation from realities:united that demonstrates the wide range of media architectural expression: Big Vortex transforms the emissions of a waste-to-energy plant into smoke rings to indicate and make people aware of the city’s carbon dioxide (CO2) emissions (see Figure 3). McCoullough (2013) states that “this is not just data visualization but also data formation,” referring to embodied information, which leads to Offenhuber and Seitingers’ (2014) conclusion that the medium – in the case of media architecture, the building – is never neutral to the viewer’s interpretation.

2Case Study: A Hybrid Media Display

In our recent work, we developed our Hybrid Media Display to address the integration of media architecture and the presentation of real-time data in the public domain (see Figure 4). Our Hybrid Media Display combines low- and high-resolution display techniques: the low-resolution screen uses light-emitting diodes (LEDs) and is capable of displaying ambient information with high aesthetic qualities. In a different mode, high-resolution information is projected on top of the LED surface to provide more detailed information about the data acting as an additional layer of annotations. We exposed the study prototype in two cases, both investigating the ambient visualization of real-time data but applied to different contexts in terms of location and content. The first case study explores the monitoring of environmental information, such as temperature and air quality, in a semi-public room. In the second case, we exposed the display to the main plaza of a residential complex, displaying the departure times of a nearby subway station.

2.4Case 1: Indoor climate

Indoor climate constantly displays real-time data on the indoor air quality with the aim of making people aware of their surrounding environments in relation to the health, comfort, and capacity for mental work of the building occupants. In the case of bad environmental conditions, people were notified through the system and encouraged to open their windows for air exchange. The determination of the indoor air quality was based on several measurement values, namely temperature, relative humidity, and carbon dioxide (CO2). These measurement values were selected because they are crucial for good air quality, most people are familiar with their meaning, and they can be measured using affordable off-the-shelf measurement hardware19.

2.4.1Location One

For the indoor climate case, the study prototype was exposed in a work room at a large university building (see Figure 5). The room was situated in the basement of the building. Thus, artificial lighting was required because not enough natural light could enter the room through the windows, leading to a homogeneous illumination irrespective of the time of day. The room was mainly used by architecture students, containing 40 work stations, each equipped with high-performance computers. Our Hybrid Media Display was placed in the middle of the window side, midway between the floor and ceiling, thus visible from all work stations (see Figure 5, top). The study location was considered promising for various reasons: The working place has high fluctuation, with many individuals visiting the area throughout the day. Thus, the ambient visualizations in the room would reach a large audience. Due to the hard-to-reach window handles and the waste heat from the computers, the bad indoor air quality in the particular room was already well known to its users. The large number of students – in particular, at the end of the term when we conducted our field investigation – intensified these problems. Thus, our application had a direct reference to the local context addressing an existing problem for which Van de Moere and Hill (2012) stress that processing “data that is sensed, measured, or acquired within the physical environment immediately surrounding the display,” is crucial for reasonable public visualizations.

2.4.2Visual Content

We designed the visual content in close alignment with the architectural design. For the pure ambient representation, three measurement values (temperature, relative humidity, and carbon dioxide) were assigned to three screens. Because of the correlation between temperature and relative humidity, it was considered reasonable to assign those values to the large, opposing faces (see Figure 5), with relative humidity on top and temperature at the bottom; therefore, the carbon dioxide was assigned to the horizontal split screen. Due to the restricted resolution, the choice of an image or text-based visualization was excluded from the onset. Therefore, we encoded the physical quantities via color: Blue represented relative humidity, red stood for temperature, and white stood for carbon dioxide levels. The measured values were encoded via area visualizations with the square screens as physical containers for the filling level. The illuminating areas represented a numeric value in data lumen representing temperature in degrees, relative humidity in percent, and carbon dioxide in parts per million. The filling level was growing toward the center of the display with increasing values functioning as a metaphor for the displacement of fresh air. In the case of temperature and relative humidity, the measured values were additionally encoded via saturation.

In order to provide more detailed information, we designed a high-resolution visualization that was projected on top of the surface. In this vein, the basic design concept for the ambient visualization was extended by symbol- and text-based explanations, scales, and divides (see Figure 5). In high-resolution mode, the filling levels with the given color encodings were adopted from the ambient visualization design. For the Hybrid Mode, using both low-resolution and high-resolution display technology, the data were presented by the LED-based filling levels and the additional supporting information, e.g., explanations and scales, projected on top as a layer of annotations.

Besides providing unobtrusive information about the surrounding environment condition, we added an alarm mode with higher notification level (Pous-mann and Stasko, 2006) to address the contextual problem of fresh air supply at the study location: If a carbon dioxide limit was reached, a visual alarm notified the users to open the windows for ventilation. After sufficient air renewal, a second visual alarm notified the users to close the windows in order to avoid low temperatures. While in the normal mode, data changes were represented subtly and slowly, we utilized fast transitions in the notification mode referring to the taxonomy for ambient displays discussed by Tomitsch et al. (2007). This setting was realized by flashing arrow-like movements on the entire display surface.

2.5Case 2: Public transport

The second case was situated in the urban realm, focusing on the ambient visualization of public transport information. The processed information was based on real-time departure times provided by the local transport company via a Web service20. The study prototype was exposed to a prominent urban site near a subway station.

2.5.1Location Two

For displaying real-time public transport data, it was necessary to find a study location in close proximity to a station. Thus, we aimed to place our prototype at a busy spot in order to provide usefulness for large audiences. We decided to conduct the second case at the Olympic Village in Munich, Germany. This popular residential district was built for the 1972 Olympic games and has 6,000 inhabitants today with a very high population density (Chalkley and Essex, 1999). The entire residential area is completely car free. We exposed our Hybrid Media Display at the main plaza of a residential complex that is primary inhabited by students. Figure 6, top shows the location of the Hybrid Media Display, the surrounding housing units, and the nearby subway station. The location of the display was approximately 3–4 minutes’ walking distance away from the subway station. The display was placed under a concrete roof construction and fastened to a central pillar (see Figure 4). Located immediately behind the covered area was a temporary used kiosk that is run and operated by students. This position was selected for several reasons: First, the covered plaza is situated in the center of the surrounding housing units and serves as a meeting point for social activities, such as celebrations or flea markets. Furthermore, the display’s position is visible from multiple spots, even from large distances. Thus, the display can be seen from a large area in front of the covered square, from the balcony of a dozen of the surrounding two-storied bungalows, and from more than 100 apartments of a high-rise flat. These zones relate to the “display spaces” – the areas from which a display can be seen – discussed by Fischer and Hornecker (2012) in their analysis of the spatial setup of media architectural interventions. In this vein, our visualization also strongly differs from existing public transport displays that are usually directly placed on the platform, not designed for long-range visibility in a city. Finally, the position of the display was also motivated by practical reasons, such as the roof construction, which protected the display prototype against rain, and the hardware, for which was possible to get power by a high-voltage system provided via the nearby infrastructure.

Overall, the public transport visualization was exposed to the Olympic Village inhabitants over a period of seven days in midsummer. Because of the strong sunlight, the projection was only visible from dusk onward; the LED Mode was apparent throughout the day. However, both reflected and emitted light sources became highly visible during the early evening up to the early morning. To protect the hardware from theft, we removed the computer and the data projector overnight after the last subway train had passed, at about 1:30 am.

2.5.2Visual Content

To obtain a better understanding what and how presented information could be useful for inhabitants and visitors in the chosen district, we conducted initial observations: While observing people who entered the subway station and talking to inhabitants of the nearby residential district where the display was placed, it became obvious that most of them mainly use the subway line in the direction toward the city center. Furthermore, it became apparent that passers-by around the plaza would consider live information about the next departure time as being the most important piece of information, as people could estimate if there is enough time left to reach the subway.

Contrary to the first use case with the high-resolution presentation strongly oriented at the low-resolution presentation, serving as a layer of annotation, in this case, the two layers were clearly separated from each other using distinct visual elements for encoding. For the projected layer, the visualization was purely text based, conveying explicit information. The presentation was derived from common passenger information displays. The line number was projected on the left side of the horizontal split screen; the remaining time in minutes with the unit symbol “min” was projected on the right side of the horizontal split screen (see Figures 4 and 6). Because the decision was made to monitor only one direction, the destination was not displayed in order to reduce the information density, focusing on a pass-by-and-use scenario (Fischer and Hornecker, 2012). For the ambient representation (see Figures 4 and 6), the remaining time was depicted as an hourglass serving as a metaphor for time. Here, the architectural design enhances the overall visual appearance, as the two horizontal screens look like the bulbs of an hourglass, encoding information in the shape of the display (Offenhuber and Seitinger, 2014). The filling levels of the hourglass were aligned with the remaining time. The filling levels were colored orange according to the color coding of the local transport system in order to recognize the subway lines. To strengthen the pictorial representation, the ambient visualization was supplemented by an animated sequence: A drop fell from the upper bulb that represented the remaining time to the lower bulb that represented the elapsed time, thus connecting both physical containers. The movement was similar to that of a water drop, which made the filling levels look like luminous liquids. Through the irregular pixel grid, the diagonal alignment of the LED panels, these slow visual animations created an additional aesthetic effect, contributing to a clear, unbounded display design. For the Hybrid Mode, using both low-resolution and high-resolution display technologies, the transport information was represented by the conjunction of the metaphoric visualization in the form of an hourglass and the previously described text-based information.

3Observations

Looking back after observing both settings for several weeks we learned that experiences of the people were very different from case to case. In the first case, at the computer room of a university, our Hybrid Media Display got ignored by most of the users of the space. This was contradicting our assumption that such an display that is a) highly visible and b) supplied with localized and situated information would provoke rich discussions and raise attention by the people populating the space. To compensate the lack of interest we tried to raise more awareness for the display by, for example talking to people in the space directly and handing out flyers which were advertising the display and its capabilities or making the animations brighter. Interestingly also these strategies did not make hardly any difference in order to raise more awareness among the students in the space. The majority of the people observed during the time the display was located there were, when entering the space, just scanning the room for a free seat and computer, carried out their work highly concentrated and left the space as soon they were done without noticing the display or asking questions about it. Conducting semi-structured interviews, we have learned that the students were simply too concerned with the work they had to deliver for exams so that they did not at all care about their surroundings or the air quality they were breathing. One student expressed that he thought the display was some kind of decoration to improve the appearance of the space and even enjoyed the animations when the display was communicating severe CO2 levels.

In the second case our installation left a very different impression as many passersby were stopping and viewing the display for longer time-spans. Because the setting in the Olympic Village was deployed during a very warm summertime week, many other people and residents were populating the square. Even during the setup of the Hybrid Media Display in this location we were constantly involved in discussions with residents and visitors. Informal conversations with architects who were involved in the design and re-design process of the Olympic Village proved to be a highlight during this setup, as the architects expressed their enthusiasm about our work: transforming their original design using mediated architecture. One resident confirmed the decorative function of the display, providing an added value for the built environment by “coloring the grey concrete buildings” and, thus, providing added value for the residential complex. Regarding the functional benefit, one occupant of the nearby high-rise mentioned that he coordinated his timely arrival at the subway station by means of the display which was visible from his apartment on the twelfth floor. Statements such as these caused us to presume that still after several days the display was perceived as an integral part of the location and the residents we have interviewed were positive about the hypothetical scenario of a permanent deployment.

Concerning the preferred display mode, interviewees perceived our Hybrid Media Display using only the LED Mode as difficult to encode the presented information. They referred to the LED Mode as being aesthetic and considered the animations as successful artistic intervention: “I found the pure LED version most aesthetic, but the most sense for the average user would make the Hybrid Mode.” In the Hybrid Mode, people were more positive and enthusiastic, considering the other modes in direct comparison: “...because I perceive the glowing LEDs being very beautiful, but more information is communicated via the Hybrid Mode compared to the LED Mode.”

The majority of the interviewed people raised further the issue of demanding more explicit information when Hybrid Media Display was in pure LED Mode. This is a common issue with ambient information systems that focus on peripheral interaction and background processing capabilities. Our chosen use cases (Case 1: indoor climate and Case 2: public transport information) demanded more precise information on time and thresholds than the low-resolution LED visualizations were capable of delivering. We consider therefore a hybrid setting using both ambient information and explicit data as an appropriate way to overcome this issue.

In the second case Hybrid Media Display served its purpose of presenting real-time data to different audiences compared to case one. Hence, one prominent insight derived from this field investigation was that the context of where media architectural installations are integrated in existing structures and surroundings play a large role. Our chosen locations were extremely different form one another in two aspects such as indoor vs. outdoor, work vs. leisure. The observation that the display received much more attention and feedback in the second case leaves us with the impression that the selection of the location and the occupancy of people in this location are factors that will have an impact on the general acceptance of any media architectural installation deployed and should therefore demand careful considerations. Hence, the busyness of people around media architectural installations affects their recognition in public environments. While work environments and highly transitory spaces as exemplified in case one might not lead to high attention when it comes to communicating ambient information, leisure environments where people can also remain for longer periods of time will likely lead to more attention and acceptance as experienced in our here discussed case study. Our findings also complement and confirm similar research in this domain as, for example, described in the chapter by Caldwell and Foth.

4Conclusion

In this chapter we presented the experience of our efforts of outsourcing information in the public domain from the focal to the peripheral attention area of people and investigated the integration of a hybrid media architecture into physical surroundings and structures.

While carrying out our preliminary observations, we could notice that our Hybrid Media Display was being perceived after several days of operation as an integral part of the study location. Considering the challenge of how media architecture displays can be integrated more seamlessly into their surroundings, we acknowledge that this can be realized rather through visual aesthetics represented by an appropriate interpretation of data that also reflects the surrounding architectural structures. Through the perception of the display as a decorative element, we have added a pleasant additional value and propose one solution of fitting ambient real-time information into the built environment.

We have also noticed that low-resolution ambient information in the public domain still requires additional explanation via pockets of explicit information. We therefore consider hybrid media architecture to be a promising means of communicating these data via ambient displays. We would like to point out that the combination of low-resolution LEDs with high-resolution frontal projection is thereby just one possible solution to overcome the issue of conveying information through ambient media architecture. Our initial insights might encourage further research, for example, on a more integrated and embedded version via dedicated areas with a higher density of LEDs or additional information layers realized using augmented reality applications to add explicit content. The latter might be an interesting solution, as people can learn and adapt to ambient data in an early stage of deployment using alternative means that can help them to grasp ambient data more easily. Proposing the aesthetic integration of data visualizations through ambient media architecture, we aim to provide an alternative solution to current city developments, and hope to enable and inspire the future work on this topic.