We have worked extensively with Barcelona’s shared bicycling system to understand and predict human mobility patterns (see some preliminary work in our UrbanSense08 paper, the presentation may be more interesting to the lay reader). This work was done in collaboration with Joachim Neumann and Nuria Oliver at Telefonica Research in Barcelona, Spain. Joachim, Nuria, and I have expanded this work significantly since the UrbanSense08 publication.
More broadly, in our research group at UW, we are interested in thinking about how existing urban infrastructures can be utilized to learn about human behavior and use that information to promote more sustainable activities. As urban infrastructures such as traffic lights, parking meters, electric meters, etc. become digitized, they provide the opportunity to be sensed and analyzed on a massive scale. This can not only reveal otherwise imperceptible qualities about human behavior, but also be used to feedback to citizens so that they have a better understanding of their cities and its underlying dynamics. As it currently stands, these Bicing visualizations tend to be visually interesting but not motivated by a specific objective. One key question for us is how can we use this dataset to promote shared bicycling on a person-to-person level (e.g., by making real-time mobile applications that tell users where the best bike route is to the next free station close to their destination, via a facebook plugin that uses social competition to drive usage, etc.) as well as how can share our findings with shared bicycling operators to optimize their systems?
Fabien Giradin made one of the first, and most famous, visualizations of the Bicing data.
From Fabien’s website:
Bicing is a community bicycle rental service in Barcelona (similar to the Vélo’v service in Lyon and Vélib’ in Paris). The stations deployed in the city offer bikes people can use for their small and medium daily routes within the city (max 30min). As part our Tracing the Visitor’s Eye project, we collected minute-by-minute data on the infrastructure status (i.e. number of available bikes for each station) over a weekend. The resulting animation shows the spatio-temporal state of the system and the mobility patterns of its users. For instance, it reveals a quantity of “cyclists†hanging out at the beach on Sunday afternoon and then returning downtown in the evening (video below).
Here is a new Bicing visualization by BCNoids, which extends Fabien’s visualizations into more sophisticated renderings such as topographical heatmaps:
BCNoids Reel from enrique soriano on Vimeo.
From Flowing Data:
Taking advantage of the data generated by Barcelona’s community bicycle program Bicing, BCNoids aims to explore the movement of 6000 bikes across a network of 400 stations. Developed in VVVV, the research draws inspiration from Craig Reynolds’ 1980s experiments with simulated flocking behviour and aspires to deliver “a tool for the analysis of human mobility patterns”.
