Social and situated mapping in Salo, Finland

Salo is well known — in Finland — as a town hard hit by globalisation — Nokia left the city — but it is also a very fragmented physical city structure. The spatial and social aspects merge into divisions between “Salo-locals”, foreigners, old people, young people, the disabled and people living in the centre and countryside. In our first interviews we have encountered a strong wish for more or better communication and possibilities for dialogue to build a new feeling of community.

Challenges: In general the fragmentation of our cities is marginalising and silencing societal groups by pushing them to the periphery. What is at stake for us here is not only the problems of inequality is “enhanced” by fragmentation, but also that this fragmentation is blocking the development of new possibilities — new ideas, networks and initiatives.

Fragmentation or abstraction / absence in various forms prevents a face to face interaction with an intuitive character that allow a much richer and deeper dialogue; where spoken and body languages are supported by a situation or atmosphere, thus enabling and opening up for something new to emerge. We believe that in spite of our advances in technology we are still in a digital and binary age that excludes vital nuances and grey zones. Where the discussion regarding the non binary is mostly linked to gender issues, we find this discussion and the critique of the binary and rational thinking relevant for our project.

About Social and situated mapping: We aim to take the present technologies and methods “to the streets” where the technology become much more ingrained into and facilitating a methodology that is situated and present in urban spaces. The idea is to use urban spaces as a more active medium in itself — doing interviews and listening to stories on site — and create an interactive feedback loop where people react and get inspired by stories they hear and thus creates new input.

We intend on working with a “Hybrid methodology” — where quantitative and qualitative methods can form a more nuanced and sensitive approach. We will experiment with both more well known “crowdsourced“ methods and compare these with what we call “curated” methods — especially a type of “street interviews” that situates the anthropological interview in urban spaces. We will deploy “SpeechBoxes” to urban spaces, allowing anyone to speak, and be spoken to. These elements are interactive, utilising technology to distill information from the spoken language

Salo Case: An important reason to choose Salo as the main case for testing methods and technology for social mapping is that we don’t risk to get seduced by the presence of a large group of (tech savvy) citizens already engaged and active. We are so to say “outside the comfort zone of the creative class”. This means that once the methods and technology is working in Salo we can be more confident that they will also work in bigger cities such as Turku and Tampere.

Tampere School of Architecture has already a collaboration with various actors in Salo with a focus on Salo High school and development of new curricula for “Civic Entrepreneurship”. This collaboration can be extended to include a broader demography by including the Vocational School in Salo. The Municipality of Salo has a wish of establishing a “Youth Parliament”. Tampere School of Architecture (TUT) will function as a coordinator that collaborates with the municipality and the local high school and vocational school in a way the young people of Salo pioneers — possibly using a game approach. The use of the “speech boxes” and the students conducting street interviews will be a way to push for a broader use that involves more ages and population groups and a as such contextualise the results in the “Youth Parliament”.

The youth parliament will be part of a longer process that gradually build up a knowledge about where and what is important in Salo. This situated knowledge will be created using both a mobile app for street interviews done by the local students and placing the speech boxes in relevant locations in the city. Both locations and topics are chosen in an exploratory manner allowing for a very open approach as a start and gradually — as a snowball effect — let patterns and dynamics emerge and inform the choices for more focused interviews. Both the concrete location and the dialogues — the agenda — of the youth parliament will be a continuation of this initial explorations.

All the field tests and explorations will have a local coordinator and also be subject of “action research” that feeds into the more strategic research being done in Tampere.

Input — processing and output

A mobile app will work as both a tool for doing street interviews and a way to communicate the situated stories on site. It will therefore have the 3 elements: Input, Data processing — same as the speech boxes — and output:

1. Input — An app that allows for doing interviews on site (and adding a number of other types of input: video, photos, maybe measurements of noise, pollution etc) and convert speech to text that is also geotagged.
2. Data processing — A database (backend) that collects the data (mainly text) and allows for analysis to discover new patterns and use this to navigate and search big data. An UI that visualise the data and allows for an advanced filtering of data.
3. Output — Same app that allows people to experience the data from the database on location using AR in various ways — maybe mostly focus on sound and listening to stories while being on location.

SpeechBox: Currently there are a number of “credit card sized” general purpose computers available on the market, the Raspberry Pi being the most famous example of the category. We will use the Raspberry Pi mini-computer as the basis of a Speech Box — an Internet-connected device which can be deployed (and survive) in city environment and which will allow anyone to speak, be heard, and be spoken to.

Speech recognition: Speech is the most natural way for humans to interact and tell stories in an urban environment, and therefore speech will be the primary, if not the only, interface to the system. Thanks to the recent progress in deep learning, high-quality, noise-resistant speech recognition is nowadays available as a low-cost service by Google.

Deep Learning: The progress in automated speech recognition and language understanding accuracy brought about through the very recent advances in deep learning makes it possible for us to aim much higher than we could have even few years ago. It is now within reach of the technology to give the voice to the people and let them express themselves freely, in a natural setting, and using spoken language. Deep learning methods are highly noise-resistant and also have proven the ability to deal with inflective languages like Finnish. We will use methods of topic modelling, text classification, and vector embeddings to relate together similar stories as well as to trace topics in time and space. The main feature of these methods is that topics can be traced regardless of the actual words used.

Summing up: Social and situated mapping is exploring new types of dialogue in and with urban spaces and their futures. Combining the newest technology — speech recognition and deep learning — with a methodology based on presence in urban spaces the project discovers and nurtures new voices of the city and urban futures. The approach can lead to innovations in what we today understand as not only urban planning or design but also journalism / communication or politics / organisation — active citizenship.