Although the sun is still shining bright from the skies over Salzburg, summer is over and the academic year about to start. Under the motto “A good start is half the battle”, our PhD researchers Christian Werner and Lucas van der Meer headed abroad to present their work, drink a pint and meet new people.
Their first stop was in Leeds to attend the GIScience conference. For both Christian and Lucas it was their first time at this event. Which is not so surprising, given that the last in-person GIScience conference was already five years ago! It brings together an interesting mix of GIScience researchers from different fields, different countries, and different career stages. Of course, advances in big spatial data analytics and GeoAI where at the forefront of discussions, but there was also strong attention paid to ethics, privacy, and equity issues in spatial data science, as well as to the limits of quantitative modelling when studying human behaviour. From our point of view, it was good to get reminded that we are not the only ones focusing on mobility research from a spatial perspective. It is a hot topic in GIScience and it was inspiring for us to discuss our ideas and findings with this community.
Lucas presented the core ideas of his upcoming PhD research on human-centric approaches to transport accessibility at the pre-conference workshop on Equitable Accessibility and Sustainable Mobility. The format with several three-minute lightning talks straight after each other was quite refreshing. Everyone had to stick to the point and was forced to only highlight the main ideas and insights of their work, followed by a large group discussion afterwards. In the afternoon the focus was shifted to open-source software tools for accessibility and mobility analysis, with several presentations and a hands-on session using the r5r and r5py packages. At the same time Christian participated in the Centrality indicators for road network analysis workshop, which perfectly matched his current PhD research. In a small but very motivated group of participants, topics ranging from application domains of centrality analysis, useful centrality metrics and their potential bias, to the location of centrality-based network analysis within the frame of a broader set of methods were discussed. Furthermore, ideas for possible extensions to the open-source routing engine Openrouteservice that enable centrality-based analyses were generated. Thank you, Christina Ludwig and Adam Rousell from Heidelberg University, for enabling this full day of knowledge exchange!
During the parallel sessions, Christian represented us by presenting his recent work on the use of spatial normalisation techniques to tackle systematic bias when computing betweenness centrality metrics in real-world spatial networks. While betweenness centrality is commonly used to assess the importance of individual roads within a transport network, this metric from network science does not consider any spatial factors. However, spatial configuration of the network (e.g., node density) has significant impact on results. Through spatial normalisation, this effect can be mitigated. If you are interested in examples, code or the associated short paper, you find all resources here. We highly appreciated the discussions on network morphology and centrality metrics that followed this presentation.
Lucas participated with a poster that introduced our development of a systematic workflow to measure bikeability in urban street networks. This is work in progress, and both a description of the current state (an open-source software package to quantify bikeability based on network properties derived from OpenStreetMap) as well as an outlook to future research directions (integrating subjective factors related to people’s perceptions of the environment they move through) were included on the poster. This was all presented in the form of a drawing, and we asked interested attendees to add their own drawings of a particular bike trip they like to do. We believe that drawings are a great way to find out what people remember from their trips and infer what they find important. For now, it was just a fun way to interact with the audience, but maybe in the future we turn it into a structured scientific experiment. Stay tuned! We were happy to see that the interactive approach of our somewhat unconventional poster was appreciated, and we got awarded the price for best poster. Thanks everyone for the participation and encouragement! If you are interested in the accompanying software (work in progress!), click here.
All in all, the “serious” part of the conference was great, but as always, the social aspects were the most memorable. We loved to see so many like-minded people coming together, said hello to some old friends, and made many new connections. A big thank you to the organisation, and we hope to be back at the next GIScience conference, wherever it may be.
Visit to ITU Copenhagen
After a weekend of exploring the stunning nature of the Yorkshire Dales and the Lake District (Christian by train and foot, Lucas by bike), the duo made a stop in Copenhagen to visit the Networks, Data & Society (NERDS) research group of the IT University. During lunchtime they presented the core directions of the research in our group, as well as the specific challenges they want to tackle in (the remainder of) their PhD research. What followed was an “extended coffee break” to discuss all kinds of challenges and possibilities in spatial network analysis, programming, but also (academic) life in general! The evening was reserved for data and beers (https://datasci.social/@DataBeersCph). Isn’t that a great combination? Interesting TED-style talks by data scientists and generous amounts of free beer. Maybe we should try to bring this concept to Salzburg as well… Thank you in particular Ane Rahbek Vierø, Anastassia Vybornova and Michael Szell for inviting us, and we hope it was the starting point of more collaborations and knowledge exchange!
In the meantime, we are all back together at our campus in Salzburg, packed with a load of new ideas and probably too little time to develop all of them. But in any case, we believe that openly sharing knowledge, discussing challenges and finding solutions together is what science is all about, and we are so happy to meet many lovely people who think the same way.