1.1Media architecture causality

For the last decade, media architecture has focused on ‘technology’ as ‘the answer’. In my 2009 publication Media Façades – History, Technology, Content (Haeusler 2009) I defined through the title a causality – understood as a relation between cause and effect – that accords Technology as the first event that is responsible for Content. The motivation for establishing this causality had its origins with the development of a taxonomy that classified the new field of media architecture. Other investigations at the time, such as the Media Architecture Conference at Central Saint Martins Innovation in September 2007 in London, also confirmed the prioritisation of technology. This is expressed by the Media Architecture Conference homepage that stated:

Developments in display technology and building materials are leading to new forms of hybrid architecture that break away from existing conceptions of surface, structure, lighting and moving imagery. Light Emitting Diode (LED) displays integrated with the fabric of built structures allow prominent imagery to be integrated with the façade, even in direct sunlight. Although energy requirements are practical only when displays are used in less bright conditions, there remain significant consequences for the townscape and urban environment; cinematic utopias as pervasive as Minority Report and Blade Runner have become technically feasible. (Mediaarchitecture.com 2007)

One could argue that media architecture’s initial pre-occupation with technology lay principally in its novelty, as by the end of the 20th century it was possible to produce large-scale RGB LED screens. In other words, the technology was new and hence offered a large area of exploration and development (Simpson 1997).

Within the first decade of the 21st century, a vast majority of economic and technological challenges of building large-scale RGB LED screens were overcome. The 2012 book New Media Facades – A global survey (Haeusler et al. 2012) documented how one could build media facades in virtually any part of the world. Here the aspect of the title ‘global survey’ suggests that Space had taken an important role as the book aimed to study cultural and regional differences in the design of media facades. I would therefore argue that after Technology, Space has become increasingly critical to the way media architecture is understood – and consequently developed by architects and designers.

Subsequently as no media façade can operate without Content, one could argue for the causality of Technology (existence of LED screens) → Space (applying LED screens in built environment)→Content (designing visual imagery) as a common way of designing media architecture in the first decade of the 21st century.

Recent debates in media architecture at conferences, summits from 2010 onwards, and in several discussions amongst the authors represented in this publication, have highlighted that this causality prevents more meaningful progress in the field. The reason for this is manifold. Firstly, the focus on technology as a driver for media architecture is in contradiction with the construction process of the AEC (Architecture Engineering Construction) industry. Most buildings that incorporate large LED screens in their building skin take years from preliminary design to completion; hence state of the art LED technology specified at the preliminary design stage is outdated by the completion date of the project. Secondly, and as a further argument against technology, is that cities, and particular those located in Asia, now host an increasing number of large screens, rendering media architecture as no longer novel but rather ubiquitous. While the sensation and spectacle of the new and novel large screen, as for example the Fremont Street Experience Las Vegas which opened in 2004 with 12.5 million LEDs, might continue to be effective in a casino city context, this is not necessarily the case in most other everyday places and cities.

Consequently, the view of many stakeholders and scholars in the media architecture field has been to ‘flip’ the accepted causality of Technology → Space → Content. This flip productively reorganizes the design process as Content (to design experiences that reflect the needs and interests of citizens) → Space (to reflect location, cultural and political parameters)→Technology (to use appropriate technologies to achieve the first two goals).

3.1Autopoietic content

To overcome these challenges a question to ask is how content is conceived and produced? Inmost cases of pre-recorded, live, participatory or interactive content, content is generated and designed by something other than itself, to be precise a designer or a team of designers, film producers, artists, etc. As already mentioned, the design of sufficient content for media architecture can be time and cost intensive, hence it is sensible to investigate other methods of design in generating a large amount of content. One approach to this, it is argued here, is to take better advantage of the already high saturation of sensors, actuators, and computers in media architecture and through sensors to transform each media façade into a machine that can generate content by itself. The two concepts of other than itself and by itself are briefly explained in the following.

In science, engineering or designing a shift from a ‘system’, ‘object’ or ‘form’ that is developed or created by something other than itself is often referred by the Greek allos (ἄλλος), meaning ‘other, different, strange’ in comparison to the Greek autós (αὐτός), meaning ‘self’. Exemplary for this use of allos or autós is an essay in architecture named Aegis hyposurface: autoplastic to allo-plastic by Mark Goulthorpe (1999) where he adds the suffix plastós (πλαστός), meaning ‘formed or molded’ to discuss a “continuity of creative process into manufacture” which “offers a precision without determinism or without auto-determinism” hence defining

Autoplastic [as] being a determinate, fixed environment – one ‘designs’, auto-dictates – and alloplastic [as] an indeterminate, open description, a reciprocal relation between environment and self. (Goulthorpe 1999)

Given this, one can argue in the case of media architecture to date, that various types of content have been primarily designed by someone other than itself (i.e. a designer) and could therefore be classified as allopoietic content, using the suffix poiesis (ποίησις), etymologically derived from the ancient term (ποιέω), which means “to make”. If on the other hand content is generated by itself, one would refer to this as autopoietic content.1

In the publication Autopoiesic content: A conceptual model for enabling situated self-generative content for public displays, Langheinrich et al. (2011) introduced the idea of autopoietic content as a conceptual model for enabling situated self-generative content for public displays. Here they traced the concept of autopoiesis back to cognitive biology where in the early 1970s two biologists, Maturana and Varela, discussed autopoiesis as a system capable of reproducing and maintaining itself (Maturana and Varela 1980). For Maturana and Varela the autopoietic machine is,

… a machine organized (defined as a unity) as a network of processes of production (transformation and destruction) of components which: (i) through their interactions and transformations continuously regenerate and realize the network of processes (relations) that produced them; and (ii) constitute it (the machine) as a concrete unity in space in which they (the components) exist by specifying the topological domain of its realization as such a network.

The term of autopoeisis was also adopted in the mid-1970s by the German sociologist Niklas Luhmann to explain aspects of his systems theory. Maturana later disagreed with Luhmann on the correct use of the term. In the context of this chapter the sociological adaptation of autopoeisis in relation to communication is considered to more closely aligned with the interests of media architecture than its origin in cognitive biology. Consequently this chapter adopts Luhmann’s reading of the term autopoiesis.

For Luhmann (1975) a system is defined by the boundary between itself and the environment. A system in its interior is reduced in its complexity; the exterior (environment) on the other hand is defined through its infinitive complex and chaotic nature. Important for Luhmann is that communication within a system operates by selecting only a limited amount of all information available in the environment, a process he calls the “reduction of complexity” (Luhmann 1982). Luhmann stresses the importance of the reduction of complexity as he associates with each system a ‘distinctive identity of the system’ that is constantly reproduced in its communication. The aspect of identity also helps to identify what is meaningful for the system and what is not. Consequently if a system fails to maintain that identity, it ceases to exist as a system and dissolves back into the environment. This process is what Luhmann defines as autopoiesis, a reproduction from elements previously filtered from an over-complex environment.

This is important to this discussion of directing media architecture towards autopoietic content generation as this ‘reduction of complexity’ is useful in (1) limiting what system we want to establish; (2) in what environment the system operates and (3) what information the system selects from its environment. Again when moving from allopoietic to autopoietic content a “reduction of complexity” and a “distinctive identity of the system” (Luhmann 1982) help to define the boundary, an aspect I will return to shortly.

Returning to the research of Memarovic, Langheinrich and Elhart (2011a, b) on autopoietic content, the authors acknowledge that their motivation for moving to an autopoietic content model follows the argument presented here concerning cost and time factors. They state that,

One of the reasons these displays are not showing customized information that resonates with their particular location and surrounding space is the high cost associated with creating tailored content (Alt et al. 2011).

They further assert that,

We believe that the idea of [such] self-sufficiency can also be used to shape future public display systems, be offering an economic way of turning today’s often ignored large public display systems into more appreciated (and more noticed) services (Langheinrich et al. 2011).

In two subsequent key papers a conceptual model for enabling situated self-generative content for public display is set out by Langheinrich et al. (2011), and initial experiences with autopoietic content are described by Memarovic et al. (2011). More specifically, Langheinrich (2011) outlines a framework for categorizing autopoietic content into three classes as:

In a follow-up paper the research team acknowledges that in developing a first autopoietic content,

The first two classes of autopoiesic content, i.e., context visualisation and context connection, have been explored to a certain extent. (Memarovic 2011)

Previously, Schmidt and Gellersen (2001) have described content visualisation examples including coloured orbs showing stock trends (ambient orb), water fountains indicating exchange rates (Datafountain), or lamps indicating web site access numbers. Yet, it is argued here that the aforementioned content visualisation examples are those that contribute to the phenomena of public display blindness as outlined by Müller et al. (2009) and Huang et al. (2008). According to Memarovic et al. (2011), people increasingly seek to see “personalised and situated content”. While it could be debated that the above mentioned content visualisations could be ‘personalised’, it is argued here that they do not consider the following two points.

Firstly, the authors state that,

This [personalised and situated content] could be provided through a menu where people could choose their preferred category in addition to choosing if content type should be local or global. (Memarovic et al. 2011)

This reference to “local or global content type” contradicts Luhmann’s (1982) understanding of autopoietic as a ‘reduction of complexity’. Equally, Luhmann associates with each system a ‘distinctive identity of the system’ that is constantly reproduced in its communication – a key point in generating an autopoietic system.

Secondly, this overlooks how public space and the surrounding building typologies already provide a distinctive identity to be used to generate autopoietic content. Memarovic et al. (2011) acknowledge that,

While some…[large public displays] are presenting contextualized information that reflects their surroundings, e.g. train schedules and university talks, a large majority is simply displaying traditional advertising, such as short commercials, slideshows, or simple images. (Memarovic et al. 2011)

Yet in their own case study of the development of the Fun Square Project, the same authors indicate they failed to present sufficient “contextualized information that reflects their surroundings” (Memarovic et al. 2011). This project instead collected local fun facts that locals associate with places or objects nearby, or in the immediate area, a point they state in the discussion of the paper as,

Although autopoiesic content connects local information from the display surrounding with information from without, some people wanted to get content that is even more connected within the locality. Some people even wanted to get localized content on specific topic, e.g., information about local politics, events, or history. With two parameters, topic and locality, categorizing content becomes difficult. (Memarovic et al. 2011)

At this point I want to point out two assumptions in order to extend the research undertaken by Memarovic et al. (2011). Firstly, media architecture and media facades are by definition part of architecture and the urban condition. Secondly, and the primary focus of the following section of this chapter, is that architecture can be defined through typology, and typologies produce the local condition in an urban setting (i.e. the business district through the ‘office’ typology or the shopping district through the ‘shops and retail’ typology). Based on this, Memarovic et al.’s (2011) problem of categorizing content via two parameters, topic and locality, can be reimagined through the concept of typology, where typology is understood to also define locality. Yet, how is typology defined in the context of the architectural discipline and what is its relationship to autopoietic content?

3.2Typology in architecture

Typology is defined as a classification according to general type. In architecture and urban design typology functions as a taxonomic classification of types. Moore and Wilson (2014) argue that

The modern concept of type has alternately been used as a method to classify both building functions – hospital, bank, house, museum, etc. – as well as building morphologies – courtyards linear, tower, block, ziggurat, etc.

The concept of architectural type has been subject of many studies spanning at least four hundred years and includes architectural scholars such as J. F. Blondel (1705–1774), Quatremère de Quincy (1775–1849), and Gottfried Semper (1803–1879). Modern typologies have been discussed and articulated as system by Nikolaus Pevsner in A History of Building Types around 1930 (Pevsner, 1979). On Pevsner’s account, Moore and Wilson (2014) argue that,

The significance of Pevsner’s insight is a major contribution to the formation of sociotechnical codes – it is to recognise that buildings are an index of changing social organisations as well as changing social tastes. Buildings are, then, technologies […] designed to assist us in our effort to make the world a better or richer place.

Further they note that

[t]he adoption of a recognised building typology by a patron is, then, to accept the social agenda of that type as a tool necessary to satisfy normal social activities… That the existence of architectural types that emerge over time – in response to changing social, ecological, and economical conditions – precedes the agency of individual designers.

Thus both Pevsner’s (1979) and Moore and Wilson’s (2014) positions support the argument that typology solves Memarovic et al.’s (2011) problem of categorizing content via the two parameters of topic and locality – as one can equal “social organisations” and “social activities” with topic and “buildings function” as a driver to the local character or locality.

Yet, how can applying the concept of typology in the context of autopoietic content for media architecture prove productive? Seen through the lens of both typology and autopoeisis, the following three observations regarding media architecture can be made:

1) Media architecture can be more than just a screen when augmented by other digital technologies such as sensors, actuators, and ICT systems (Haeusler 2009, 2011).

2) When equipping building skins with screens and digital technologies one can argue that a building equals an autonomous system, able to sense and collect data. This data can then be processed into information to communicate to other buildings or humans to generate new knowledge about the building itself or the urban context (Haeusler 2011).

In the case of media architecture currently, communication occurs mainly from the building to humans, but not yet from a building to a building. Yet if systems become increasingly autonomous and buildings can be autonomous systems – where buildings communicate with buildings – the argument here is that their dialogs can constitute autopoietic content. The implementation of this autopoietic strategy for media architecture, however, requires addressing further questions. In order to communicate information between buildings, buildings need not only to know what data to collect, and how to process this data into productive information. More importantly, a building should know and understand what another building is interested in and what a building does not need or want to know. To further outline this point, consider an example from a human context, presented in a simplified story taken out of a situation in daily life. Imagine a social gathering where a soccer fan and a fine arts lover want to discuss and communicate their interests to each other. Each party may seek to stimulate interest in their specific subject from the other party, yet they remain talking about their specific focus, perhaps obstructing the potential for a longer conversation. The reason for this, ignoring here for a moment communication theoretical concerns and argumentations, is that there is no common topic of interest to discuss. Being aware that this statement might imply a entire set of assumptions regarding the structure of human communication, for simplifying the argument we can argue that if the fine art lover is substituted with another soccer fan, and even if both soccer fans support different teams, a conversation can occur through the common foundation of soccer itself.

Consider now the example of the fine art lover and soccer fan as buildings that are differentiated by typology, the fine art lover is an office building and the soccer fan is a train station. While it seems reasonable to assume two train stations can exchange information, to what advantage, and further, how can an office building ‘talk’ to a train station? Things that are relevant for train stations such as timetable updates, passenger numbers, ticket information and so on, may or may not be of value to the office building. Yet, between various buildings of the same typology, such as two train stations, or a bus stop and a train station (akin to the two soccer fans who support different teams), common topics can be readily established and productive interaction is made more feasible. This brings us to the third and last observation.

3) Communication, and rules about language and the use of language, has been argued by William Allman (1995) to be relevant to the formation of a civil society. Given this, if buildings can communicate, so it follows that they require a set of rules – not to generate a society, but rather to establish and maintain a self-generating system. Two possible rules for communication between buildings are considered here.

In a city system, and based on the previously discussed topic of typologies – non-typology specific rules and typology specific rules can define an interaction between systems.2As argued prior with Moore and Wilson (2014) “buildings are an index of changing social organisations as well as changing social tastes” These social organisations define what users expect from a building (in transport environments the notion that the user will be transported from A to B and is informed about how to start his/her journey from A to B via a set of information such as customer information [what mode of transport]; wayfinding information [where does my journey start]; time table information [when does my journey start] and ticketing information [how much is my journey]). Hence it is not the typology as such that has communication capacities but the social organisation that can be communicated from one type (a bus stop or taxi station) to another similar type as all types within the same typology provide the same “index of changing social organisations as well as changing social tastes”. In outlining the difference between the two sets of rules two comparison cases are discussed in the following to show the difference between non-typology specific rules and typology specific rules.

Non-typology specific rules can, for example, describe the potential communication between the typology of an office building and an apartment block. The second typology specific rules can describe the communication within a typology, such as transport. The first, non-typology specific rule, can be compared, as discussed in the chapter on observations, with a conversation between a soccer fan (i.e. office building typology) and a fine art lover (i.e. apartment block typology) where both have very clear rules (passion) within their system (hobby), but one has to first establish and understand rules (a common discussion point) amongst both systems to achieve a holistic phenomenon (a long lasting conversation). The second, typology specific, on the other hand is similar to two gardeners (i.e. bus stops) who have both the same interest (rules of interaction amongst parts) talking to each other about horticulture (timetable information and bus arrivals).

Furthermore, it is argued that non-typological and typology-specific rules can form the basis for generating autopoietic content as they follow Luhmann’s (1982) argument of “reduction of complexity” and a “distinctive identity of the system”.

To illustrate this point further the following section discusses a research example of a typology specific rule set applied within the context of a public transportation system.

3.3‘Bus Stop of the Future’

In the context of public transport infrastructure research (train station, bus stops, ferry terminals), the concept of typology has been operatively adopted in the following research example to achieve a reduction of complexity and to make the social, economic and operational demands of public transport more comprehensible and manageable. This follows the argument that – as Gardner et al. (2010) argued on the topic of responsive environments in Infostructure: A Transport Research Project – the pressure on public transport systems can be addressed in three key ways: by improving the proximate access to a transport service (physical and contextual), by improving access to service frequency (operational and demand led), and by facilitating and managing better access to information (commuter interface). In discussing typology in the context of autopoietic content the research has favoured and followed the last point on improving access to information, and to be precise, enabling this through the application of digital technologies.

This research topic has been the focus of an Australian Research Council project funded since 2010. This has culminated in the production of a mock-up bus stop as a test set-up for future autopoietic content development. While the scope of this chapter does not allow for an extensive outline of the development process, the publication Interchanging – Future Designs for Responsive Transport Environments (Gardner et al. 2014) further details the conceptual framework employed in the development of the bus stop mock-up. Furthermore, the chapter entitled, ‘Responsive Transport Environments – System thinking as a method to combine media architecture into a digital ecology to improve public transport’ in the book Digital Futures and the City of Today – New Technologies and Physical Spaces (Clift; Smith; Caldwell 2016) frames this approach from a theoretical perspective.

In short the developed bus stop made use of a range of number of ICT components such as screens; Mac Minis; Microsoft Kinect sensor; LEDs and control systems; and WiFi; Bluetooth, etc. Intentionally, all components were purchased from a large local electronics supplier, and subsequently nothing ‘exotic’ was specified for the project. This was an important point for the team as we aimed to demonstrate that Urban ICT does not require expensive state of the art technology, but rather can be achieved with ‘everyday’ consumer electronics.

The bus stop design concept was developed in January 2014 at the University of New South Wales as part of a design studio for 3rd and 4th Bachelor students from Architecture, Computational Design, Interior Architecture, Landscape Architecture, Industrial Design and Construction Management. In parallel, Masters of Design Computing students at the University of Sydney developed a series of apps to be displayed on the screens of the 1:1 prototype.

With the bus stop’s design finalized the team commenced production using the facilities available at each of the universities. Based on the optimization phase, the team fabricated the bus stop out of 18mm plywood panels using a 3-axis CNC mill.

Ideas and concepts for apps were based on shifts in the day-to-day engagement with digital technology, and how such apps might better influence how we commute and travel. A key objective and challenge was to translate interaction design concepts that work on a personal device with a small-scale screen, to a public device used by potentially many users concurrently on a larger-scale screen. In detail these were: real-time bus timetable and seat availability information (displayed on the large LED screen); active surveillance system (displayed on the 60′′ screen facing inward) and local business information (displayed on the 42′′ screen facing inwards) (see Figures 1–3).

Notably, the design of the hardware and the software had a key research constraint worth mentioning here and elsewhere in other publications before. As there was only one bus stop prototype, and as this mock-up was situated, first in a museum, and secondly outdoor on university campus, testing and evaluating of apps and hardware was not possible and constraints and findings will be published shortly.