The adaptive city: A socio-technical interaction-driven approach towards urban systems

PROCEEDINGS OF THE INTERNATIONAL

BiENNiAi

CONFERENCE

HYBRID CITY 2 0 1 3

SUBTLE REVOLUTIONS

ciTy

I

NJVERSITY RESEARCH INSTITUTE OF APPLIED COMMUNICATION

D . CHARITOS

I . THEONA

D . D R A G O N A

H . R i z o p o u L O S

M . M E I M A R I S

^EDITORS^

SUBTLE REVOLUTIONS

Proceedings of the

l""^ International Hybrid City Conference

Athens, 23 - 25 May 2013

National and Kapodistrian University of Athens

Edited by:

Dimitris Charitos

louliani Theona

Daphne Dragona

Charalampos Rizopoulos

Michael Meimaris

The Adaptive City

A socio-technical interaction-driven approach towards urban systems

Achilleas Psyllidis

Faculty of Architecture, The Why Factory & Hyperbody

Delft University of Technology (TU Delft) Delft, The Netherlands

A.Psyllidis@tudelft.nl

N i m i s h Biloria

Faculty of Architecture, Hyperbody Delft University of Technology (TU Delft)

Delft, The Netherlands N.M.Biloria@tudelft.nl

Abstract.Jhis paper aims at establishing an associative relation between pervasive digital technologies, the physicality of the urban fabric and its inhabitants. It introduces a methodological framework for the development of an interactive urban system, installed within urban open public spaces, in the form of a hybrid interface that can serve as an interactive platform for both citizens and local planning authorities. This particular system apart from harnessing and visualizing real-time diverse quantifiable data, derived from everyday urban activities, aims to further provide the inhabitants with an agency via a continuous feedback loop processes to, ultimately, influence the physical and behavioural patterns of the city. In other words, the platform does not only imply interaction at an information exchange level, but rather aims to provoke a complex variety of inter-relations between the social and the technological via real-time spatial adaptation and spatial customization possibilities. The proposal focuses towards a system that is perceived as an integral part of the urban environment and less on the development of a specialized application or website platform that only overlays an additional virtual layer to the already existing ones in contemporary cities. Lastly, the paper deploys a set of critical issues that need to be taken into account regarding the evaluation of such systems in practice.

Keywords: Urban systems; Interaction design; Urban computing; Real-time systems; Adaptive urbanism; Ambient interface

1. INTRODUCTION

The perception of cities as complex systems has been widely argued in the last two decades [1]. Urban complexity appears, though, to increasingly proliferate, triggered mainly by two pressing issues; on the one hand the rapid urbanization processes and, on the other, the perpetual pervasiveness of information technologies within the urban environments. The first contributes to

the increase of complexity/density on the physical fabric of the city, while the latter augments it with a multiplicity of virtual layers in an unprecedented manner. Nevertheless, these two significant aspects of contemporary urbanity need not be treated as separate.

This researcli is co-financed via tlie act "Scholarsliips Program of the Greek State Scholarships Foundation (I.K.Y.) with individual evaluation procedure acad. Year 2011 - 2012" by the resources of the Educational Program "Education and Lifelong Learning" of the European Social Fund (ESF) and the NSRF, of 2007 - 2013. It is also funded by the A.S. Onassis Foundation, as well as by the Foundation for Education and European Culture (I.P.E.P).

but as important parameters of an emergent hybrid urban geography.

Global urbanization processes have exceeded the milestone of the balance between rural and urban populations. Already since 2007, more than half of the world's population lives in cities and, according to UN predictions, by 2050 it is estimated that this percentage will rise up to 70% (United Nations, 2007). As a result, humans are already - and increasingly become - an "urban species" [2]. China represents a prime example of such procedures, since it is currently halfway into a major urbanization process, according to which by 2020 four hundred cities of one million inhabitants each are -and continue to be - built from scratch [3]. This can be translated in the creation of about twenty new cities per year, i f we consider the fact that the program was announced officially in 2001. In other words, what has to be comprehended and accounted for, is the transition of an amount of people that equals to more than half of Europe's population from rural towards an urban state. Yet, these cifies are still being planned within a top-down master-planning framework, which supersedes the emergent values of the 21st century. This rapid urbanization process fiirther necessitates that the responsible design community, which gives shape to the inter-relationships between the social activities and the urban context, start devising systemic interactive processes, which provoke trans-scalar interrelations between the social and the technological; namely in an era where information becomes more and more pervasive.

A direct consequence of the perpetual technological ubiquity can be identified in the obvious shift from an industry-based economy to one driven by (digital) information and services, which we have experienced primarily during the last decade. Though the concrete repercussions of digital media on the city have been quite controversial [4], we can firstly acknowledge a clear influence of pervasive and situated technologies as regards the amounts of tangible patterns and traces of human urban activities. Industry-driven societies have been characterized by a plethora of visible activity patterns in the physical spaces of the city. Manchester,

for example, the world's first industrial city - also niclmamed "Cottonopolis" - has been largely depicted as a dense aggregation of factory buildings with protruding chimneys in an endless production of smoke. Further, the proliferation of job opportunities in the industrial city centres, as well as the advances in the infrastructural sector have led to the well-known gigantism of the urban configurations. A l l these illustrate typical cases of the direct impact that the technological progress has upon the physical scale of the city. On the other hand, the growing diffiasion of ambient technologies within the urban environments has gradually led to the invisibility of the "by-products" of human activity. The physical repercussions of the service- and information-based work become, on a first level, less and less traceable. Yet, the layout of the contemporary, digitally augmented cities appears not to be different from that of the industrial ones. It is primarily this reason that the aforementioned controversy is based upon, as regards the impact of pervasive media as opposed to the shape of the urban fabric. However, we could claim that the vast amounts of virtual data produced simultaneously on a daily basis by diverse sources influence namely the way we experience the city. They represent the city's "pulse" [5], which figuratively appears as an additional, intangible layer hovering above the urban fabric, contributing to the emergence of a hybrid reality of both physical and virmal. Respectively, as inhabitants of cities we are provided with the opportunity to become aware of several aspects embedded within the sun'ounding urban environment in an unprecedented way, primarily with the assistance of sophisticated devices or webpages.

We argue, though, that this particular perception of ICT media and the derivative ambient data as a superimposed layer over the existing fabric, implicitly suggests that information is incapable of drastically affecting the urban layout. Such an overlaid ontology presupposes the dominance of the built components over the informational ones, rather than the latter being part of or equal to them. Within this framework, the paper challenges the idea for a model of urban systems in which the diverse amounts of data derived by social activities become equally co-constimting with the physical environments they belong to. Not hovering above the city, but embedded within it. We propose a framework - currentiy in progress as part of an ongoing PhD research conducted by the authors' - for a complex urban system plugged into existing open-air public spaces, which receives data as real-time streams and acts upon these meta-data. The fiinctionality of such a system can be, more precisely, described in a twofold manner; on the one hand, it hamesses data in the form of quantifiable, real-time digital traces of urban

' The working title of the PhD research is "The Adaptive City: Research on the Configuration of Real-Time Interactive Urban Systems", conducted by AchilleasPsyllidis, under the supervision of Prof Winy Maas (The Why Factory - MVRDV) and Assist. Prof Dr. NimishBiloria (HyperBody), at Delft University of Technology, Faculty of Architecture.

activities (e.g. occupancy levels, transport and mobility patterns, energy data) and, on the other, it feeds them back in a looped manner to both citizens and local planning authorities, via public interfaces, providing them the opportunity to develop an active dialogue upon diverse datasets, with qualitative outcomes. By devising such a systemic interactive urban meta-system we aim to further provoke more intricate inter-relations between people, activities, context and technologies, appropriate for contemporary and fiiture cities of the 21st century. In this way the physicality of urban space, assisted by the proposed hybrid system, might become more connected to the human side and, ultimately, raise the potential of changing local behaviours, social attitudes or activity patterns that - in a networked manner - will increase the liveability of urban public spaces on a global city scale.

I I . T H E C I T Y AS A COMPLEX A D A P T I V E SYSTEM Prior to deploying the functionality framework of the proposed urban system, it is important to comprehend what potential does the theory of complex systems provide to our approach. In different words, we need to fiirther elaborate with what we mean precisely by the term "urban system". Cities themselves are also perceived as complex systems. The idea is not new, since it falls under the field o f Complexity Theories of Cities (CTC), which already counts three decades within the urban planning and design discourse [1]. Yet, many of these theories provided urban simulation models that were largely applied in a mechanistic rather than an adaptive way to the city. The issue of adaptation is, though, an important aspect of complex systems and, i f applied properly, it can act as a steppingstone towards the development of complex adaptive systems within urban agglomerations. Within this consideration, the following paragraphs provide an overview of the systemic approach of cities.

A. Complex vs. Simple Systems

In general, a system - derived from the Greek word moTfjfia (systèma) - can be characterized as a set of inter-related, yet autonomous elements. According to Harvey, a system can be more accurately defined as a set of elements with certain variable characteristics (attributes), along with a set of relations between these element-attributes (which constitutes the structure of the system), as well as a set of relations between the element-attributes and the environment [6]. It is this particular characteristic that diversifies an open from a closed system. Open systems interact with their environments while closed ones do not. It is, though, important to be noted that closed systems are not existent in reality; they only constitute human-made abstractions, used namely for the simplification of aspects that are characterized by high complexity levels. Complex systems, including cities, are necessarily open, meaning that all their constituent elements along with their corresponding attributes can change in time owing to the inter-relations developed between the different elements as well as between the system they constitute as a whole and its environment. Within this context, our aim is to perceive the proposed urban system

embedded within the physical fabric - as a complex one. It is not only fed raw data derived from a multitude of parameters constituting the city itself (the immediate environment) but, it in-tum feeds these data back in a looped manner, while engaging local inhabitants as well as experts (system - environment interaction) and thus ultimately results in adaptation and customization of physical space. In this way, it forms a constituent part of a wider city system, while referring to it at the same time. Thus, we can perceive it as a form of meta-system.

B. The Emergent Value of Adaptation in Cities

The notions of adaptation and adaptability are critical issues of complexity theory. Adaptation, originally developed within the field of evolutionary biology, refers more specifically to the structural, fiinctional or behavioural change of living organisms over time, so that they are able to survive in a specific environment [7]. Huxley, further, categorizes adaptations into three features: structural, behavioural and physiological adaptations [8]. The first refer to specific body elements that facilitate and allow the organism to survive in its natural environment (e.g. body covering, shape, skin colour and texture). Behavioural adaptations concem the specialized ways in which an organism behaves in order to fit in its natural habitat (e.g. phototropism). Finally, physiological adaptations are special systems embedded in an organism that allow it to perform certain biochemical reactions (e.g. secreting slime, homeostasis).

On the other hand, adaptability is a critical property of complex systems, stemming from their capability of openness and self-organization. This particular attribute enables complex systems to behave as complex adaptive systems (CAS), which means that they are capable of adapting their stractural, behavioural and physiological features to their suiTounding environment [9].

In the urbanism domain, adaptation - though still on a nascent level - can formulate an emergent value for ground-up, responsive urban systems. Their components can be thought of as ubiquitous entities embedded in the larger city-"organism", comparable to swarm elements or natural cells, which are endowed with local intelligence akin to a bold, while autonomous adaptations of these cells in co-ordination with their immediate neighbours lead to different kinds of adaptation. In our case, we are specifically interested in the behavioural changes that the proposed urban system would trigger, firstly on a local scale of a certain impact radius and, secondly, on a global city scale through the networked connection of the different sub-systems. C. Real- Time Relational Model of Urban Data:

Towards Complex Adaptive Urban Systems

The challenging concept of implementing adaptive urban systems should, though, keep us aware of avoiding the pitfall of rendering cities equal to natural organisms. Cities are by no means natural systems, but rather completely artificial ones and they need to be perceived and treated as such. However, what urbanism can learn from adaptation processes is to shift focus shift from a predetermined form to the relations between

the constiment elements that configure the complex city system. In other words, it challenges the development of symbiotic systems based on real-time urban data, as a form of a "new geography", that provoke more inli-icate inter-relations between people, space and ambient technologies. This, fiirther, implies the necessity for a relational model of the diverse patterns of uitan activities. Unlike the current prevailing approach of treating various quantifiable activity traces (e.g. occupancy levels, transport and mobility patterns, energy data etc) as individual systems, it is important to begin smdying the impact of these traces as a result of the interdependent relations that can be established between the diverse infrastructures and people, in a more sustainable and interactive manner. In this way, we believe that the city will start performing as an organism that better adapts to its environment and caters to the needs of its inhabitants. We can thus essentially envision complex adaptive urban systems that enhance the efficiency and services of the city, in which they are embedded.

I I I . T H E FRAMEWORK OF THE A M B I E N T A D A P T I V E U R B A N M E T A - S Y S T E M

The rationale of the proposed ambient urban platform can be broken down into three sub-categories, which provide a general image of its fiinctionality: Urban data sensing and gathering, Indexed data simulation and visualization and, ultimately. Feedback loops via public interfaces that can drive real-time infrastructural adaptations within a certain impact radius.

A. Urban Data Sensing

As already mentioned, the basic principle of the proposed urban system is not premised on an overlaid ontology, but rather on its perception as a flindamental constituent part of the environment it inhabits, so that it allows more direct and tangible human-machine interactions. In different terms, within the city as a complex system we intend to integrate another dynamic system, as a plug-in in existing open-air public spaces, which is fed directly by data derived from its immediate environment. To achieve this, on a first level, the platform incorporates a physical sensing system consisting of low-cost sensors, which are distributed across a certain study area and are based on open-source hardware, such as Arduino and Servo motors. Where possible, existing open datasets from dedicated govemmental databases will be harnessed, to further facilitate the real-time data mining process and, also, keep the cost level as low as possible. Through these procedures, the goal is to create a relational model of the indexed datasets, so that the impact on the urban fabric, derived from quantifiable measurements, will not only respond to individual parameters, but would rather refer to the repercussions resulting from a relation that can be established between different elements and attributes of the city (e.g. between people and traffic levels, between people and environmental conditions etc). However, as cities have become highly complex and not even two open-air public spaces or neighbourhoods, within the city, are identical, the

proposed system is not correspondingly intended to work upon generic components, applicable almost everywhere. Instead, it operates as an open-ended system that can engulf variable quantifiable attributes, so that it adapts itself to its immediate environment issues. It should, thus, give the opportunity to plug in different parameters that address emergent matters, otherwise the system will soon become obsolete.

The data hamessing process aims, additionally, at the reinforcement of the socio-technical role of the proposed urban system. Therefore, the identification of user-centric variables will be utilized for the deployment of a PACT (People, Activities, Context, Technologies) analysis-driven design framework, to extract the proposed interactive interface requirements (Fig. 1). This is of major importance, since we are aiming at a case specific urban system. The PACT framework creates a synergy of the four components by establishing a single statement; people undertake activities in (certain) contexts using technologies. It, thus, qualifies as a logical concept to sustainably study the interdependent relations between the four components. Urban data stemming from occupancy levels of a specific public space, or demographics can cmcially provide an insight into the operational components, especially those referring to people and their activities in the case study areas. Though the exact elaborations as regards this design framework fall out of the intentions of this paper, what is important to comprehend is the cyclic nature of interactivity among the four components, which is vital to the attainment of a successful urban system.

B. Data Simulation and Visualization

After the data hamessing process, the system simulates, analyses and visualizes the digital activity traces, with the assistance of customized scripted routines. This procedure is not, of course, developed in a linear way, but rather in an iterative manner, where urban data are constantly fed into the software platform in real-time and the visualized outcomes are in mm communicated via public interfaces, which will be fiirther elaborated in the following section.

As was the case with distributed sensing hardware, the visualization and analysis processes take place

Requirements - V Activities within Context s X .

; people ;

Fig. 1. Tiie People, Activities, Context, Technologies (PACT) desien framework's rationale.

within the open-source Processing programming language and integrated development environment, which is based on Java. The procedure is fiirther facilitated by specialized libraries that extend the potential of the basic Processing application programming interface (API). The choice for this particular programming platform has been made mainly due to its open-source nature and free distribution, which significantly reduces the total operational costs of the urban systems. Further, it comprises a rather flexible programming language that can easily manage a variety of added parameters, thus meeting the criteria for the configuration of an open-ended system. In our research, specifically, we are mostly interested in particular datasets that refer to occupancy levels in different time phases of the day, transport and mobility patterns as well as real-time energy data (e.g. CO2 emissions, electricity usage etc). These groups of data are relatively easy to be quantified and tracked, either via distributed sensors, RFID tags and mobile phone signals or accessed via dedicated open governmental/municipality databases. This last issue is vital, since currently a great amount of different datasets is only accessible by a very restricted number of people or authorities. O f course, a city comprises numerous diverse components that equally produce a variety of datasets, which all together influence our experience within the urban fabric. Yet, it is almost impossible to integrate such a multiplicity within the limited space of a research framework. Thus, only a glimpse of carefully selected parameters is examined as well as their impact within a specific radius that, i f scaled up on a wider level, can provide us with an image of potential adaptive processes on the city scale are dealt with. In turn, through quantifiable measurements and illustrating the processed deductions in an easy-to-comprehend fashion, we expect qualitative feedback from the citizens themselves who integrally co-constitute the perpetual variety of information sets.

In terms of simulating the human activity pattems within the smdy areas, a swarm-behaviour model is applied, based upon the principles of C. Reynolds' model [10]. Such a model depicts, in general, the behavioural aspects of a group of agents that may be able to perform tasks without detailed representations of their immediate environment, as well as other neighbouring agents [7]. Originally, the model as devised by Reynolds was meant to simulate complex natural systems, such as floclcs of birds, schools of fish or animal herds - by developing the simplified flocking creatures, loiown as boids - through the use of three basic steering behaviours; separation, alignment and cohesion. A l l three are based on the positions and velocities of an individual boid's nearby flock mates. In an urbanism context, the apphcation of swarm logic can provide an interesting simulation model, especially because it endows its component units with a sort of agency, unlike other computational models such as cellular automata (CA), in which every single monad's - or cell's - properties are solely determined by its neighbouring cell states. Though, it should be noted that despite the apparent ability of the swarm simulation model to cater to the more intricate relations developed between the urban agents, we need to be aware of the

potential as well as of the restrictions that the model implies. As shown in related research, Reynolds' model appears to be highly successful in terms of namral systems, but it might not be as equally sufficient as regards human aggregations [11] & [12]. Consequently, when such a model is applied in an urban context, it needs to adapt to the artificial nature of the city. However, what we can benefit from the application of swarm logic, as a simulation tool to urbanism, is the potential to explore organizational behaviours that evolve in time, derived fi-om localized simple rules of action. Thus, the diverse measured urban components can be, on the one hand, visualized as real-time data processing entities and, on the other, the model can give us the oppoitunity to study them in a relational manner, which - as already mentioned - comprises a focal point of our approach towards adaptive urban systems. On a different level, swarm simulation can be extremely usefiil for what has been described as "artificial planning experience" [13]. This can be of a major importance, especially for the part of local planning authorities, which also play a role in the proposed interactive system. With the assistance of this model, local planners can be provided with an artificial experience of the impact that certain interventions can have within a specific area, in a relatively short amount of time. For instance, when testing possible scenarios they can project - almost in real-time - flimre usages that can affect activity and occupancy pattems in a certain radius of built space. Yet, this has to be proved also physically by, for example, observing how patterns are changing (with the help of the distributed sensor system) in certain proportions, which is a work in progress and exceeds the intentions of this paper.

Apart from the technical requirements and models used as regards the urban data visualization and simulation procedures, another critical issue refers to the form of outcome, in which the processed activity traces will be communicated with the citizens. It needs to be presented in a clear and easily accessible way, so that it engages larger groups of citizens to actively evaluate and comprehend the interdependent relations between different components of their immediate environment, as well as stimulate them to dynamically participate in the qualitative feedback process conceming their spatial and infrastmcmral usage alternatives. To increase the potential engagement of citizens that are not familiar with sophisticated computational environments, data visualizations in our proposed urban system are - as already pointed - part of the system's functionality, thus physically embedded in the urban enviromnent that creates this information. We intend to avoid the communication of the processed meta-data via solely , context-specific webpages or mobile phone applications since, as aptiy observed by Vande Moere& Hill [2], they are separated from the environment that constimtes the data source and, thus, transform the urban experience into a virtual one or, as we remarked, they turn it into an overlaid ontology.

C. The Public Interactive Interfaces

While the process of data visualization is important for the comprehension of the impact of urban activities

in the immediate environment, what constimtes the key characteristic of the proposed ambient urban system -and, subsequently, the major challenge of the research in general - is to provide an interface, as a form of interactive platform for citizens as well as municipal planning authorities. The challenge is, thus, not to solely provide the citizens with visualizations of a high aesthetic value, but rather enable them as active agents that can observe, interact and declare their own activity-driven customized spatial and infrastmctural usage and transformation alternatives. In this way, the system provokes a constant feedback loop process, from data harnessing to visual representations and simulations of processed meta-data, from interactive response upon the meta-data, to real-time spatio-temporal and service-related adaptations and back again, in a recursive manner. Within such a framework, city inhabitants can actually get involved in the transformation process, rather than barely get notified of various information sets.

This section speculates - since it refers to a work in progress - on the potential characteristics and requirements that the interface should fiilfil, in order to facilitate real-time adaptations on the physical activity patterns of the city. It fiirther proposes a possible scenario, which showcases the urban system's operational processes towards a hybrid socio-technical urban spatiality.Thus, it should nol be conceived as a given, but mostly as a methodological framework that outlines ideas and principles for possible implementation in the contemporary or the near-future city. In particular, the platfonn's fundamental goal is to provoke more intricate relations between people, their activities, the immediate urban context and technologies. To achieve this, it should allow more complex human-machine interactions, which subsequently imply the necessity for tangible interfaces with physical presence in the actual urban fabric. Even though the very concept and vision of ubiquitous computing, as devised by Mark Weiser [14], favours the idea of invisibility - a property that is currently supported by various smartphone devices, RFID tagging systems, GPS trackers, as well as by what is referred to as "the Intemet of Things" - it is cmcial for our system to make people aware of its existence, so that it leads to a higher potential of human engagement. Of course, in terms of information processing - as regards the data gathering, simulation and visualization procedures - the principle of pervasiveness, along with the subsequent invisibility, is highly legitimate. The outcome, though, of such processes as well as the ability to actively participate in the aforementioned feedback loops, should be integrated in physical installations, simated within the urban environment. It is even supported that by making the constitoents of an urban system and their in-between interactions visible to the public, we will achieve a fiirther understanding on how the system operates and, thus, familiarize ourselves with it [15]. Insofar as digital information becomes a visible, constiment element of city's system it would substantially influence our experiences and potentially lead the urban space and services to adapt to the citizens' activity pattems.

Nonetheless, given the current nascence of urban computing as a research and practice field, along with the restricted amount of data sources available to public access, there is a limited amount of public displays that we can, presently, propose as potentially successflil examples of interfaces triggering physical space adaptability. Moreover, it should also be noted that our research would substantially benefit from the outcomes of the PACT analysis framework, which will provide us with more case-specific requirements for the proposed urban platforms. As of now, we are aiming at a hybrid configuration of urban screen displays providing visualized representations of situated information and, most importantly, giving the city inhabitants the ability to feed the system back with their spatial, infrastmctural and transformation alternatives. These public displays will be distiibuted in open-air public spaces, but need to move beyond the logics of advertising banners and announcement boards. As it is akeady observed in related research, the combination of data hamessing either via open-access sources or via distributed sensor networks, along with feedback provided by citizens can result even to real-time alternative perspectives of the city [16]&[17]. In our case, we are currently studying our proposal's alternatives in the real-world urban context of Rotterdam city centre.

More specifically, the proposed public interface displays consist of the processed visualizations stemming from the locally harnessed datasets, as well as consist a user-friendly platform that can accept citizens' feedback on the meta-data. Municipal planning authorities also play an important role in the system's process. Particularly, they evaluate the simulated outcomes from the swarm behavioural model, in order to propose potential interventions in specific parts of the city for increasing their liveability and efficiency. These proposals can also be displayed in the interactive platforms, so that the city inhabitants are able to express their opinions and, possibly, provide fmitflil remarks as regards their empirical experiences of the city. Hence, the local planning authorities are able to receive feedback information in real-time and initiate strategic decision-making procedures. Yet, such a process will lead to medium or long-term adaptations of the physical fabric. But the interactive platform is also capable to cater to immediate demands and act upon real-time adaptations, especially in terms of infrastructural or service operations. Since urban activity is monitored and instrumented by a distributed network of sensors, the information that is gathered and processed in the software platform can, ultimately, have a real-time impact on service operations within a certain radius.

More precisely, a potential operational scenario o f the proposed meta-system could incorporate a variety of floor pressure and proximity sensors that can be embedded within urban squares (for our case in Rotterdam) to measure occupancy levels in real-time throughout the day. This network of sensors can be linked to the already existing ones, currently measuring transport and mobility pattems, as well as with sensors dedicated to weather and energy data (i.e. air quality.

CO2 emissions, electricity consumption levels).All the information harnessed by the distributed sensors consists the data input to the urban system. Subsequently, via specific algorithmic routines, the system produces visualized outcomes pertaining to the indexed data. These visual pattems of the ubiquitous, amorphous and un-processed information derived from the sensor network, constitute the system's meta-data that are presented onto the public displays/interfaces. Both citizens and municipal authorities can have access to them and further act upon.

On the one hand, local planning authorities are able to draw meaningflil conclusions upon the extracted pattems. According to the occupancy levels, they can evaluate possible latent public spaces within the urban fabric. Based upon this, they can fiarther proceed to a multiplicity of more sophisticated actions. For instance, bus or tramlines can be re-routed in real time to where it is mostly demanded, so that it efficiently caters to the needs in a specific area of the city. At a next level, with the help of the previously mentioned simulation outcomes (swarm-based), planners can expérience the repercussions on a specific area of the city as a result of the transportation changes, pertaining to the relation between occupancy levels and environment-related parameters. By combining the visualized and simulated after effects, they can initiate possible intervention scenarios that adapt to the conditions of a certain area within the city. These alternatives can, then, be loaded to the system, so that they become visible and widely accessible via the public interfaces.

On the other hand, citizens can actively participate in the urban environment configuration process. Apart from the proposed interventions, they are provided with customization possibilities within a certain framework that also adjusts to the available budget of each specific case (e.g. increase the amount of trees or green surfaces, select through a variety of materials/colours/shapes of urban furniture or installations, change in real time the light levels/colour in a specific public space etc). Thus, the form of meta-data that the citizens have access to and can act upon is two-fold: firstly, they can obtain a real-time image, for instance, of which squares currently host a small amount of people and which are quite crowded. Subsequentiy, they can choose directly which one to visit according to their mood. In addition, the system can provide them with information on the quickest possible route and means of transport that leads them to their choice, as derived from the transportation and mobility data. Further, they are aware of the current microclimatic conditions at the public space of their choice (e.g. i f there is sufficient amount of shadow during the warm days, humidity levels etc). In this way, they will be given the opportunity to combine information in an unprecedented relational manner. Secondly, as previously mentioned, they can act upon professional intervention proposals and contribute to their customization. To be more democratic, the incorporation of the customized suggestions in the intervention proposals will follow the average amounts regarding each given customization possibility. In turn.

the informed decisions of people can lead to the mobility patterns' differentiation, which subsequently has an impact on its linlced parameters, so that they constitute a continuous feedback loop.

This further implies that, by utilizing the same urban information system, various adaptations can occur in the physical urban environment within different time spans. Time is, thus, a cmcial parameter of the system's rationale. But what is more important is the system's ability to facilitate inter-connections between the different time spans. Hence, the decision-making process, for example, and the subsequent interventions in the urban fabric, as proposed by the municipal planning authorities, can result into long-term adaptations but, on the other hand, these proposals will be facilitated by feedback information derived in real time from the city inhabitants, as well as from simulated models of urban activity pattems. Thus, apart from constructing productive and provocative trans-scalar inter-relations between the social and the technological, the system also assists in configuring simultaneous connections between adaptations taking place in different time spans.

I V . DISCUSSION : O N THE CHARACTERISTICS OF INTERACTIVE U R B A N SYSTEMS

With regard to the aforementioned concepts and potential practical examples, this section synopsizes some of the critical characteristics that need to be taken into account in the development process of the proposed urban system. Yet, by exemplifying upon the following issues, we aim at a fiirther contribution to the wider discussion about the potential influence of complex urban systems on our everyday experiences in the contemporary and near-future city. Obviously, the requirements and characteristics described below do not comprise a comprehensive overview, but rather an open-ended framework for current and fiimre perspectives withm the nascent research field of urban computing and informatics.

The foremost principle, which our urban system is founded upon, refers to its incorporation in the physical urban environment. The system needs to be embedded and have a physical presence, so that it, on the one hand, provides a tangible platform fof the city inhabitants and, on the other, allows for a hybrid interlocking reality between the physical and the digital to emerge. In direct relation to this notion, the identification of the variables and information processed by the system are closely related to context-specific characteristics and the corresponding human activities, taking place in the area where it is embedded. The diverse datasets can originate,

from different open-access repositories (municipal, web-based, sensor and mobile phone signals), to better provoke a synergy between the social, the spatial and the technological. Respectively, any technological changes affect the manner in which the activities within a certain context are performed. This necessitates the open-endedness of the system's model, so that it caters to continuously changing requirements, stemming from alternating activities.

Nevertheless, for the system to i n f l „ .

effectively its immediate urb'an

J L ^ ^ ^ ^

pursue suggestions as well as collect and evaluate data within a certain radius of the city. Subsequent hi a dec entralized networked manner, the interlinked be al platforms can have an impact on the global citv scale bringing about emergent and unpredictable outcomes' Yet, especially as far as the measured variables are concerned, it is important that the visualized representations resuhing from them would provoke the presentation of multiple relations - which are often not obvious - that can be established between the different components, so that they are consistent with the perception of the city as a complex system.

Another critical issue that the urban system needs to comply with is that of interactivity. Being solely responsive - in the way that, for example, an automated door currently operates - is not adequate enough. The system has to interact with different users, either city inhabitants or municipal authorities or local planners, meaning that it has to go beyond a linear, cause and effect relation between input and output, towards a more intricate system of inter-relations between people and technologies [18]. Subsequently, wider groups of people - o f different ages and cultural backgrounds - will possibly get engaged with the system and actively participate in the formation of their urban environment. Ideally, in the course of time, the system - provided that it is augmented with a sort of intelligence - can actually leam from everyday urban activities and even suggest possible alternatives, adapting to frequent usage patterns. Or, even further, in the near fiiture the system can move beyond merely interactive towards "transactional" stmctures, as mentioned by Adam Greenfield, in which each interacting party exchanges something of utilitarian value with each other [19],

However, in order for all the aforementioned issues to be tackled, the urban system has to engage different parties in making active use of it. Especially when the interfaces involve public displays, specific requirements need to be fiilfilled, such as display positioning, size and content dynamics, to increase the possibilities of attracting human attention [20].

V . CONCLUSION

Contemporary cities are encountered with unprecedented global urbanization processes, as well as with a proliferation of ambient information technologies. However, our current urban configurations are hardly any different from the ones established during the Industrial Revolution. But even though the layout of our cities appears quite static, the transition to an information-based economy has instigated an emergence of increasing inter-relationships between people, context, activhies and technologies within urban space. These processes necessitate that the responsible design community, which gives shape to such inter-relationships starts devising systemic interactive-planning processes, which provoke trans-scalar links between the social and the technological. The Adaptive City proposal, within the aforementioned context, intends to interactively transform the physical

urban fabric as an interface to digital infonnation; Infoimation that is derived from social dynamics, in terms of a symbiotic "new geography" of people, space and ambient technologies. It, specifically, proposes an interactive urban meta-system, which harnesses and analyses digital traces of urban activities (e.g. occupancy levels, ti'ansport and mobility patterns, energy data) and subsequently feeds this analysed information back in a looped manner to citizens via public interfaces where they can observe, interact and declare their own activity driven customized spatial and infrastmctural usage and transformation alternatives.

We argued about the significance of physical presence within the urban fabric as a fundamental attribute of such a system, so that it provokes a hybrid interlocking between the immaterial flows of infonnation and the material space of the city. In this way, the increasing amount of data produced daily by diverse urban activities will be fed back to the environment that produced them, instigating civic action with direct impact in the physical space and infrastmcmres, rather than simply remaining on a virtual level solely accessible via certain mobile phones or websites. The challenge is, thus, to provide interfaces that endow citizens with a sort of agency, rather than just giving them back bare data mappings. Nevertheless, it is crucial that municipal planning authorities also operate these interactive systems, in order to provide certain frameworks and boundaries within which citizens can declare their altematives. Subsequently, our proposal does not aim at an absolute bottom-up or top-down decision-making strategy for the city, but instead favours a merger of both professional proposals and everyday people's suggestions, as an emergent hybrid approach towards urban design. Fumre work will focus more on exemplifying upon the specific ways in which this merge can be successfully attained, as well as upon the technical requirements and characteristics of the system's soft- and hardware and the subsequent repercussions it can provoke in the physical urban environment. The Adaptive City proposal, thus, aims at enhancing the already existing hybridity of the urban fabric, by developing the framework for a hybrid socio-technical urban system, facilitated by hybrid decision-maldng strategies, as a sustainable approach for the twenty-first cenmry city.

REFERENCES

[1] J. Portugali/'Complexity theories of cities: achievements, criticism and potentials", in ]. Portugal!, E. Stolk, E. Tan [ed.], "Complexity Theories of Cities Have Come of Age", Heidelberg: Springer, 2011,pp. 47 - 62.

[2] A. Vande fvloere, D. Hill, "Designing for the situated and public visualization of urban data" in Journal of Urban Technology, vol. 19:2, pp. 25-46, April 2012.

[3] N. Mars, A. Hornsby, "The Chinese Dream: A society under construction", Rotterdam, The Netherlands: 010 Publishers, 2008.

[4] A. Picon, "Toward a City of Events: Digital media and urbanity", in New Geographies, vol. 0, 2008, pp. 32-43. [5] N. Leach, "Swarm Urbanism" in Architectural Design (AD),

vol. 79:4, pp. 56 - 63, July/August 2009.

[6] D. Harvey, "Explanation in geography", London: Edward Arnold, 1973.

[7] N. Biloria, "Interactive morphologies: An investigation into integrated nodal networks and embedded computation processes for developing real-time responsive spatial systems" in Frontiers of Architectural Research, vol. 1:3, 2012, pp. 259-271.

[8] J. Huxley, "Evolution: The modern synthesis" (3rd ed.], London: Macmillan Pub Co, 1975.

[9] J. H. Holland, "Hidden Order: How adaptation builds complexity", New York: Helix Books, 1995.

[10] C. W. Reynolds, "Flocks, Herds and Schools: A distributed behavioral model", paper presented at the ACM SIGGRAF '87, published in Computer Graphics, vol. 21:4, 1987, pp. 25¬ 34.

[11] S. R. Musse, C. Babski, T. Capin, D. Thalmann, Daniel,"Crowd modelling in collaborative virtual environments", paper presented at the ACM VRST '98, Taiwan, 1998.

[12] C. Loscos, D. Marchal, A. Meyer, "Intuitive crowd behaviour in dense urban environments using local laws", paper presented at the Theory and Practice of Computer Graphics, 2003, pp. 122-129.

[13] ]. Portugali, "Self-Organization and the City", Heidelberg: Springer-Verlag, 2000.

[14] M. Weiser, "The Computer for the 2Ïst century", in Scientific American, 1991, pp. 94-104.

[15] A. Powell, "Wi-Fi, resistance, and making infrastructure visible", in B. Crow, M. Longford & K. Sawchuk (Eds.], "The Wireless Spectrum: The politics, practices and poetics of mobile media", Toronto, Canada: University of Toronto Press, 2010, pp. 172-186.

[16] A. Vaccari, F. Calabrese, B. Liu, C. Rad:i, "Towards the SocioScope: An information system for the study of social dynamics through digital traces", paper presented at The 17th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, pp. 52-61. [17] K. Kloeckl, 0. Senn, C. Ratti, "Enabling the Real-Time City:

LIVE Singapore!", injournal of Urban Technology, vol. 19:2, pp. 89-112, April 2012.

[18] H. Dubberly, U. Haque, P. Pangaro, "What is Interaction? Are there different types?", in Interactions, vol. XVI.1(69-7], 2009.

[19] A. Greenfield, M. Shepard, "Urban computing and its discontents", in 0. Khan, T. Scholz, M. Shepard (eds.],'Architecture and Situated Technologies", Pamphlet 1, New York: The Architectural League of New York, 2007. [20] E. M. Huang, A. Koster, ]. Borchers, "Overcoming

Assumptions and Uncovering Practices: When does the public really look at public displays?", paper presented at the International Conference on Pervasive Computing, Berlin: Springer, 2009, pp. 228-243.