“What kinds of new territories of participation are opened up when cycles of everyday action and the representation of collective sovereignty are bound so much more closely within planetary information networks, now responsive to a molecular level? Does representative democracy evolve into a democracy of representation, and if so, what does that mean when every inch of the world comes online, becomes awake to express its informational existence to us and for us? Would that expression come as a din of voices we are incapable of listening to, let alone governing through: a churning cacophony of signals?
(Bratton, 2009)
A Novel Urban Experience
Since 2007, year of production of the first IPhone, the diffusion of smartphones has sustained an accelerated production of large repositories of data that constitutes today the basis of an increasingly augmented urban experience. Following a general tendency to multi-functionality and miniaturization, our mobiles have in fact developed from bare sound transmission systems into personal interfaces for the planetary-scale computation.
While the expanding sensoring functionalities – including the ability to collect data related to movement, acceleration, orientation and proximity to the human body – trace new possibilities for the depiction of intimate collective behavioural maps, the offer for continuous online connectivity has opened up novel opportunities for the production of situated urban diaries, that in the form of micro-blogs, social networks and reviews, constitute an emerging source of qualitative urban intelligence.
As these new privately generated sets of data become available, some of the resulting issues related to privacy and ownership are being addressed by an increasing number of providers through circular mechanisms of exchange that see users trading personal information for free services. Google Navigator is only one of the many online resources that make use of this logic to gain real-time urban intelligence, with his customers trading geo-local information against traffic updates generated through the collective movement feeds.
This new and wide-spreading attitude to data-sharing can also be read more in general in connection with the increasing popularity of Open Data ideas, also testified by the ratification of the many EU and US directives that have sustained the recent creation of multiple governmental platforms for the download of raw public data for general analysis and use.
Despite fears of data misuse or privacy invasion, it seems in conclusion that the return of real-time informational services has provided a fundamental edge in generating wide-spread consensus regarding the use of anonymized behavioural data, ultimately sustaining the booming popularity of the many online tools that are already reshaping the way we use, navigate and engage with our urban environments.
Within this context, it’s unmistakable how a contemporary approach to the exploration of urban dynamics increasingly has to rely on the understating of the complex interactions between physical spaces and the traces of our daily online activities.
And while contemporary urban formations together with the underlying data-spaces emerge as novel forms of synthetic organisations, it is in their very peculiar nature, intertwining granular behavioural and geo-spatial information, that designers must search new occasions for harvesting social and collective intelligence, and to establish novel scenarios of public engagement in the making of urban space.
Harvesting Crowd Intelligence
Nevertheless, and despite the fact that an increasing number of archives are being made publicly available through direct download or API interfaces, the task of extracting actionable information from raw data remains beyond the abilities of the untrained.
It is mainly in the field of marketing research that a wealth of collection, analysis and visualization methods have been developed to extract crowdsourced intelligence from online scraped resources. Moving on from classical statistical modelling these techniques are increasingly relying on machine learning algorithms to develop predictive analysis of quantitative data to develop business insights and support smart decision-making.
But quantitative data are not the only focus, with text based sentiment analysis gaining momentum as method for extracting strategic and actionable intelligence from online resources. “We Feel Fine”, a project developed by artist Jonathan Harris and computer scientist Sep Kamvar in 2006, is perhaps a prototypical example of how of qualitative data depicted from textual information might serve the purpose (Kamvar and Harris, 2011).
Conceived as an emotional search engine, the project run as online platform from 2006 till 2009, crawling blogs and social networking sites to extract sentences including the words “I feel” or “I am feeling”, as well as the gender, age, and location of the people authoring those sentences, and allowing users to run searches or browse over the results, asking questions about how did people of certain age or in a certain location felt about specific situations or events. Installed as interactive web-based art piece, the project was ground-breaking for its focus on providing flexible data collection and as an engaging medium of qualitative exploration for emotional data.
Similarly, some of the work developed by my students at Kent School of Architecture, where with Mattia Gambardella I run MArch Unit 4, explore the use of online data-mining for urban analysis and design and the related issues of actionability of quantitative and qualitative resources within this context.
In particular, Unit 4 looked this year at the harsh and inhospitable landscape of the Reykjanes Peninsula in Iceland, investigating the island’s rich cultural and natural heritage and the emergent environmental issues, in connection to ideas of adaptability and transitional land occupation, to develop alternative models of hospitality in response to the booming demand for touristic offer in the area.
The students developed interventions at the urban and meso-scale, initially considering the peninsula through advanced data-visualization, linking online data retrieval with geographical information for the creation of granular perceptual and notional maps. Ground conditions, modes of transport, social network feeds and image archives were collected scraping governmental agencies platforms, community websites, photographic libraries and social networks, to create individual frameworks for systemic and responsive design proposals.
The results of this initial investigations reveal a series of engaging maps that by interactively overlaying granular geo-spatial and qualitative information not only depict how both inhabitants and tourists inhabit, move and engage with the peninsula, but offer detailed insights by age, gender, profession or interests. These maps also constitute the basis for the definition of strategically distributed touristic offers. Further, some of the students were also successful in devising specific behavioural agencies and deploying them on their chosen site to simulate the patterns of access and navigation ultimately informing the special organisational and structural logic of their architectural proposals.
While Unit 4 recent research offer some relevant insights on how architectural and urban design might become responsive to qualitative emotional agencies, it also clearly identifies the issue of scalability of territorial and urban data as the central area of concern in the mechanism of their activation. Not only in fact urban data must embody spatial information, but the quality of such information must be granular.
Actionable and Playful
Breathe! is FLOW Architecture’s proposal for the Tallinn Architecture Biennale 2017 Pavilion, awarded the 3rd prize in the International Competition. The practice, which I lead with Vincent Nowak, pursues a distinctive digital research agenda exploring the role of crowdsourced intelligence within urban design and fabrication methods, to which the project contributes with a proposal for a series of collectively manufactured urban bio-filters for the purification of air in the pedestrian areas adjacent to the Estonian Centre of Architecture.
According to the European Environment Agency (EEA, 2017), air pollution by airborne particles represents a concerning issue in Tallinn, where airborne particles, mainly produced by vehicle engines, road abrasion and construction sites, retain a high level of toxicity, particularly in the trafficked areas around the city centre, where also the Estonian Centre of Architecture is located. The widespread concerns have sustained in recent years policies of rigorous air monitoring, with the key air contaminants being continuously audited and shared with the wider public in the form of real-time diffusion maps through the website of the Estonian Environmental Research Centre (EKUK, 2017)
But, despite the availability of granular information, general remediation strategies have lacked to address effectively the condition of pedestrian commuters crossing the most trafficked areas at peak times. FLOW’s project proposes to involve TAB’s visitors in the activation of these environmental data-sets, asking them to participate in the monitoring and analysis of the air quality on site. The proposal provides a mechanically ventilated voxel-based timber structure, designed to support the growth of perennial creepers, exploiting their ability to absorb the airborne particles and the gasses produced by the passing vehicles.
By scanning one of the QR codes at the base of the pavilion, the visitors are invited to preview on their phones a detailed diffusion map that describes the real-time distribution of airborne particles on site, before placing on the timber frame an individual nutri-sticker designed to direct the growth of the perennial creepers towards the areas more likely to be confronted by high levels of pollutants.
Guided by the spreading arrangement of the acid green stickers the timber frame is gradually contaminated by the growing creepers, with its final arrangement slowly emerging at the end of the Biennale as a response to the collective interpretative effort of the visitors. The result is an additive manufacturing strategy that, by mapping the history of optimisation of the pavilion as urban bio-filter, explores the aesthetics of embodied behavioural patterns within the public domain and the added value of playfulness as engaging mechanism for distributed decision-making.
In conclusion, despite the pieces of work discussed above represent very initial explorations of how online social behavioural patterns might be used in the context of urban design, they also present evidence of how a systemic design approach might be able to define responses to emotional agencies, embodying collective intelligence at the very granular level of the building proposals, and ultimately initiating novel forms of public engagement for design and manufacturing.
We have reviewed how the process of actionability of urban archives shall encompass the three steps of collection, representation and analysis, but also how urban data must embody granular geo-spatial information to allow for scalability and responsiveness at the level of the architectural proposal. While Breathe! tackle the issue by setting out a project-specific collection process, the work of KSA Unit 4 offers insights of possible alternatives with the attempt of reproducing perspective behavioural patterns at different scales and locations through agent based behavioural simulations.
Collectively these pieces of work also display how the process of engagement might surface at different points of the actionability chain, and how each step – collection, representation and analysis – might offer the potential for different feedback loops on the granular data. In particular, by directly exploiting the choreographed act of collective response to the environmental maps as actual and progressive manufacturing process, Breathe! makes use of the specific setting of the Biennale to explore original modes of engagement at the representation and analysis levels of the chain.
But if the work presented here certainly do not trace a conclusive picture of how emotional intelligence and behavioural patterns might be able to inform performative urban design, while offering significant insights on how the research might develop further through behavioural patterns simulation and the exploration of the deriving granular engagement mechanisms, it also describes how systemic architectural practice might lead innovation in urban decision-making and participatory processes by developing engaging novel aesthetics.
References
- Bratton, B. (2009). In Bratton, B. and Jeremijenko, N. (2009) ‘Suspicious images, latent interfaces’, in Situated Technologies Pamphlets. New York: Architectural League of New York, p
- EKUK (2017). Estonian Environmental Research Centre. [online]. http://airviro.klab.ee/modelling/en. [Accessed 31/01/2017]
- EEA (2017). Air Pollution – State and impact (Estonia. [online]. http://www.eea.europa.eu/soer/countries/ee/air-pollution-state-and-impacts-estonia. [Accessed 31/01/2017]
- Kamvar, S. D. and Harris, J. (2011) ‘We feel fine and searching the emotional web’, Proceedings of the fourth ACM international conference on Web search and data mining – WSDM ’11, p. 117. doi: 10.1145/1935826.1935854.