Idea & Cooperation Partners

The CityLAB invited four universities to explore the potential of data-driven innovation together. The goals were developing new insights and tools as well as a better understanding of how our cities need to change and adapt in the face of modern challenges.

The participating parties were:

The Institute for Urban Futures at the University of Applied Sciences Potsdam

The Einstein Center Digital Future in cooperation with the Technical University Berlin

The Department for Information Technology, Communication and Economy at the HTW Berlin

The Human-Centered Computing Lab (HCC)  at the FU Berlin

Data Partners

Partner organizations, like Deutsche Bahn, Technologiestiftung Berlin, ProPotsdam, Bikesharing providers like Nextbike, BVG and others, have provided data sets along with lectures on challenges that the cities across the globe are faced with at the onset of the 21st century.

Teamwork

After the topics and data sets had been presented, everyone was free to decide which topic he or she wanted to participate in. The team that gathered around „Bikesharing“ consisted of: Clement Lefevre, Leo Hohenbild, Damla Çay, Anna Meide, Felix Jaekel, Julian Schiemann.

Since we came from different countries as well as from different disciplines, an intercultural and interdisciplinary team emerged.

We decided on Telegram Messenger for communication, Google Drive as organisational Platform and Figma for design work, since those platforms can span over different operating systems and allow for simultaneous workflow.

Topic & Approach

The relevance of urban, environmentally sound mobility derives from the fact that urban populations are increasingly growing. Combined with global challenges ranging from digitization to climate change this means that cities require innovative ideas and knowledgeable implementation of the big data, that is at their disposal. We asked ourselves – how could this be done using bike sharing data?

Dataset

The dataset consists of mobility data from two major bike sharing providers (LIDL-Bike and Nextbike) in the urban Berlin area. The dataset covers the period from April to Mid-July 2019. It includes stationary as well as free floating bikes. The data points we use are the start/end timestamp and location of each ride. From this, we extrapolated the corresponding duration and distance. This dataset was provided by the Technologiestiftung Berlin, namely by Alexandra Kapp and Fabian Dinklage to whom we would like to express our gratitude.

Aim of the project

The aim of the “Bike Pulse“ project was to make the data set usable for the city. The final objective was to enable policy makers in charge of the city’s infrastructure to make informed decisions driven by adequate data visualization and analysis.

Data Visualization

Practical workshops and comprehensive introductions to various data sets were provided by the CityLAB throughout the whole time.

After a quick exploratory data analysis (EDA), we aggregated the data along the geolocation and time interval (morning, noon, evening, night). With resulting data, we used the programming language R and the Uber Deck.GL visualization tool to build a proof-of-concept (POC).

We started out by looking at the data to understand what questions we could and could not ask it. For example, if we would not have gotten any time related data, then it would have made no sense to build a concept around let’s say duration.

In the course of the analysis also the question arose as to which form of visualization best conveys which statement. In other words – what is the most suitable form of visualization for a particular statement?

Interface

After having created several data visualizations with the help of deck.gl, we turned our attention to the question of how we could use them to the maximum benefit of the city. It took several ideation phases for data-driven solutions before our final concept took shape.

The result was a simple prototype of a real-time interactive data display to monitor the biking behaviour within the city.

The display, designed for a typical desktop, should be easy to use for users who are not data experts, who are afraid to deal with big data and who want to encourage and support big data in their decision-making process. With strong colours and a dark background we wanted to give the display a stylish look and increase the joy of use. Big data should not only be associated with effort and complexity, but also with important and cool potential.

Poster & Exhibition

The format of the final presentation was a group exhibition. We presented our data visualizations on a desktop monitor, accompanied by a poster explaining the project.

Evaluation

Several important insights have emerged from the work on the project.

It is important what questions are asked of the data. Because data does not speak for itself, but must be questioned guided by an interest in insight. Data that is not integrated into a meaningful concept remains unused and data should not be an end in itself.

When processing data, not only the type of data visualization is important, but also the interaction. Access options such as zooming, filtering, comparing, etc. enable a deeper and better understanding of the data and hence the situation at hand.

We are certain that an interactive dashboard with a user friendly interface can enable decision makers without technical background to use it comfortably and quickly get an insight. In doing so they will get a better understanding of what cities need and how to change and adapt them in the face of modern challenges.

Big data in the realm of urban mobility for example should definitely be made open access for the city administration by big players like DB, BVG, car and bike sharing providers like Nextbike, navigation apps providers like Google and alike.

Limitations

The data sets we used as basis of this project do not provide the itinerary of each trip, thus limiting the possibility of a deeper analysis of the bike mobility pattern. Furthermore, as bike ownership in Berlin is historically higher compared to cities like London or New York, the present datasets do not necessarily reflect the general biking patterns.

If there had been more time, I would have been very happy to develop a click prototype and do user testing in a real environment, such as a town hall for example.

Reflexion

By working on this project I understood my role as an interface designer much clearer.

For interface designers it is immanently important not only to know about web layout but also to have a basic understanding of information architecture and how to convey an idea. In addition, interaction possibilities with the machine or data are the heart of an interface.

From the communication point of view, it is also about mediating, e.g. between data experts and city employees, because what appears crystal clear to one group and supposedly needs no mention is often not the case for the others involved.

I am sincerely grateful for for this hands-on experience and know that they will make me a better interface designer down the road.