From September 26 to October 5 at the Medialab Prado, Visualizar questioned the use of urban open data. Feedback from the developer sent by Ping and the Makery medialab, Julien Paris.
A ten-day hackathon punctuated by conferences, as many experts and professionals available to facilitate and assist developers, and a work environment, the Medialab Prado, very comfortable and teeming with life. The association from Nantes, Ping and the Makery medialab sent the developer Julien Paris to take part in the workshop Visualizar 2016: open cities, as part of their collaboration in the field of data visualization.
The best way to enter the Fabcity, of which Paris will be the epicenter in 2018. The Fablab Barcelona, supporting the Fabcity initiative, took advantage of Visualizar to organize a long session around the Fabcity Dashboard, a dashboard to observe the impact of the movement of fablabs involved in the relocation of production. Julien Paris did take part and reports for Makery.
Visualizar, a dataviz hackathon
For its ninth edition, the theme of the Medialab Prado Visualizar program was urban open data (summarized in the subtitle “open cities”). The lecture style or theoretical style conferences and the hands-on demonstrations of applications allowing you to manipulate and visualize data were held to punctuate an ongoing workshop lead throughout the event.
After ten intense days, the participants from Spain, Argentina, Mexico, the United-States, Denmark, and France, from diverse backgrounds (designers, architects, coders…), had to present their results for the opening of the Open Cities Summit.
Theories and practices of the dataviz
“Informing is not necessarily knowing.” It is with this warning not to get lost in the meandering of technical possibilities that Gonzalo Abril Curto from the Complutense university in Madrid opened Visualizar 2016. In this journey through the history of visualization going back to Galilee’s diagrams and based on the readings of Walter Benjamin, Gonzalo Abril makes a clean distinction between speech modes and data production modes. Narration is a speech mode produced using traditional methods, the novel is a manufactured product. Information is a product of industrial capitalism. The parallel between the extraction of raw material and data mining therefore makes perfect sense.
Ignasi Alcalde (Open Data Institute Barcelona) proposed a typology of several forms of visualization. Infographics we come across in newspapers, static and informative, is the most common of them. It only connects a little the information to the user’s knowledge. On the opposite, the dynamic dataviz enlightens the user and pushes him to explore the data himself. Between these two extremes, Ignasi Alcade places hybrid forms, animated and commented infographics, such as the video Wealth Inequality in America.
“Wealth Inequality in America”, Think Reality infographics, 2012:
María Poveda Villalón, doctor in artificial intelligence at the Ontology Engineering Group of the Universidad Politécnica de Madrid, invited people to tutorials where each group attempted to make a drawing of the ontologies of their own project. A very unsettling exercise when, as in the case of the Fabcity Dashboard project, the sources are multiple and the collected data is of very different nature.
Then came the most thrilling presentation in my opinion, that of Propublica, by the journalist Sisi Wei, who made us discover certain subjects of investigation data-journalism processed by Propublica, American media with the NGO status that is celebrating its 8 years of existence. Its economic model based on donations allows it to exist without resorting to advertising. Propublica does investigation journalism, which means that its writers can spend two to three months delving deeper into a subject, call on the community of readers to process or collect data.
Aside from theoretical considerations, the multiplicity of presentations and citizen projects developed in workshops gave an insight of the scope of resources available for those who are interested in open data. Non exhaustive inventory.
Javier Perez presented the application Tipi Ciudadanos, a tool to follow the development of laws voted at the Spanish congress, the aim of which is to develop a transparency culture of governmental institutions as well as encouraging citizens to participate more and beforehand to the process of writing laws. By proposing a thematic approach to access the different corpus and warning tools, the founders of Tipi Ciudadanos wanted to make everyone’s access easier. The application acts as a kind of scanner of official publications and also calculates a certain number of statistics reflecting public interest or Parliament activity.
The Stadatus project, an audit portal for public policies, scrutinizes the law on transparency, access to public information and the good governance voted at the Spanish parliament in 2013. Do national and municipal services (the municipality of Alicante serving as an object of study for the prototype developed in Madrid) truly enforce the law, and are they making efforts in order to comply with it? Beyond the “supervision tool” aspect of public policies, Stadatus is considered as a space where citizens could afterwards propose additional criteria (such as transparency indicators based on international indicators) in order to reinforce the openness of public data.
Liquen is a platform project where would be collected the environmental data of a city that calls on crowdsourcing to provide feedback such as noise pollution, cleanness. Part of the development focused on a mobile version to facilitate the data-citizen’s work that project leaders call for. An online version should see the light of day soon.
The group gathered around issues of public health (“Social Salud / Abierto directorio de recursos”) thought of proposing access from a unique platform to all public services of the Madrid metropolis, by creating thematic entries according to needs and users’ areas of interest.
The platform Apps4citizens presented its catalogue that unites several hundred applications (Spanish) aiming to involve citizens in the life of the city on a local scale. The objective is to improve the redirection of Internet users towards applications geographically close to them by inventing a system that automatically geo-locates catalogue entries.
Several proposals around mapping were presented, including Vizzuality, a company that maps matters of public concern such as deforestation, depletion of resources or still energy consumption. We were also able to test CartoDB and Quadrigram, two applications offering true computing power for the analysis and access to public data–but that gave me the impression of being labyrinthine systems for punctual use.
The impact of fablabs on resiliency
The lab movement means at the same time a tendency of these places to promote the relocation of production activities within cities, a change in consumption and distribution habits, but also new ways of considering work. The Fabcity Dashboard project—a dashboard for Fabcity to which I participated (I’ll get back to it more in details)—sets the hypothesis that such a movement can help cities face future changes due to climatic changes, economic crisis or demographic evolution. It offered to make an application prototype where all available indicators could be visualized in order to gauge how the presence of labs can foster the resilience of a city.
The purpose of the workshop was to come to a first visualization tool (some of them being cartographic) where one could cross-check and analyze existing open data on labs and indicators related to global, national or urban developments. The Fabcity Dashboard team was composed of Massimo Menichinelli (project owner at Barcelona Fablab), Marianna Quintero (Fablab Barcelona) and myself.
Ten days of workshop were far from enough to result in a functional application but they were necessary to grasp the size of the project: define concepts such as fabcity or urban resilience; identify and examine sources for open data; select relevant indicators at this stage; develop a web application prototype and its technical aspects; define one or more narration modes for the application so that it matches different user profiles… and prepare a presentation in view of the opening of the Open Cities Summit! If there is no finished product, the basis for collaborative work to come is set.
The data homogenization glitch
As Greg Bloom pointed out in his Openreferral presentation, and as we experimented with Fabcity Dashboard, the more open the databases, the more difficult it is to assemble them and make them dialogue with one another.
The official institutions (city halls, governments, international institutions) work differently, from one institution to the other. Civil society organizations (NGOs, associations) also have their own logics. We have dug into the databases from the OECD, the European Union, into open source geographic data, or even into the open data portals of several large cities and we have noticed that it is difficult to homogenize the headings in order to prepare cross-check analysis.
We have chosen to write scripts that retrieve data via the APIs of the source sites and homogenize indexes within the application. Nonetheless, there is a general problem regarding the lack of standards for open data.
Visualizar has gathered the 2016 project documentation on Github
Access the Fabcity Dashboard prototype source code