People’s mobility, and thus bicycling, is spatial by its very nature. Being mobile by bicycle means to ride from one location to another in a given environment. Fundamental geographical characteristics, such as neighborhood, accessibility or distance, determine mobility to a certain decree. However, these interdependencies are often neglected in bicycling research, planning and politics. The consequences of non-spatial approaches become evident in many cities: the environment (neighborhood) of bicycle ways is not considered and thus often unattractive or not suitable, central facilities are poorly linked to bicycle infrastructure (accessibility) or not straightly connected (distance). The graphs below show the increasing distance travelled by commuters in Austria. The proximity between place of residence and workplace directly affects the mode choice.
In order to explicitly consider the spatial nature of bicycling mobility and to relate multiple perspectives on the environment, Geographical Information Systems (GIS) are increasingly employed in bicycling research and promotion. GI systems are capable to model and digitally represent all relevant physical objects (road infrastructure, facilities, land use etc.) and moving subjects, including quantitative and qualitative attributes. Using the geographical coordinate as common denominator, all entities, together with domain-specific attribution can be related to each other. This way, additional insights and new information about the multifaceted system of bicycling mobility can be gained.
Such integrated approaches are especially beneficial in the context of bicycling, where not only rational, but also subjective (for example with regard to safety) factors, together with interests of various stakeholders need to be considered. Facing and adequately addressing this complexity is also relevant in bicycle-related research. Explicitly geo-spatial approaches leverage existing domain knowledge and contribute to better results. Representing, modeling, analyzing and visualizing different perspectives on bicycling in a spatial framework leads to new knowledge and a strong evidence-base for informed discussions, participation processes and policies.
At this year’s POLIS conference I’ll present three case studies, which proof the integrative power of geography and the contribution of GIScience to bicycling research:
To strengthen active forms of mobility, it is necessary to adapt the road network in a way which allows optimal usage in spatial as well as temporal respect. The research project FamoS , started in September 2016, investigates the potential of traditional demand based traffic models (“4-step-model”) and of agent based simulation models to estimate the volume of bicycle traffic for entire cities at a maximum detailed scale level. These models are then fed into a novel planning tool, which facilitates evidence-based decisions in the process of planning and (re) organizing public space for active mobility.
The research project GISMO , started in October 2016, integrates domain-specific know-how from various disciplines, namely GIScience, sports medicine and mobility management. As part of the project, the health effects of several interventions that promote sustainable, active mobility are investigated in a clinical study. These data are then combined with spatial models and analysis routines in a comprehensive map-based information platform, where the spatial characteristics of commuting trips and expected health effects are considered in mobility recommendations on an individual level. For a brief project update see my last post here .
3. Planning a Bike Sharing System
In order to transfer existing knowledge on Bike Sharing Systems (BSS) and parameters to a specific urban setting and to provide an evidence base for decision makers, we applied a generic spatial framework to the city of Salzburg (Austria), which merges spatial analysis results, expert knowledge and feedback from citizen participation processes. With this approach the potential demand could have been estimated for any location in town. Moreover, the contribution of each station location to the entire system was spatially modeled and optimized.
The spatial framework will be published and presented at next year’s TRA conference in Vienna.
In all thress presented cases solutions emerged that would have not be possible in the respective domain silos. However, the geographical space (concepts from geography and GIScience) is an efficient facilitator for cross-domain collaboration and knowledge generation. Domains (such as health science and medicine) and applications (such as transport modeling) which are often disconnected from bicycling research and promotion are integrated on the basis of common geographical coordinates. Consequently, the complexity of bicycling mobility can be better addressed when various perspectives on bicycling and respective interdependencies are explicitly considered.
P.S.: The presentation is available on Slideshare
Started in fall 2016, the ongoing research project GISMO (Geographical Information Support for Healthy Mobility) is the first of its kind – at least here in Austria. It brings together domain expertise from very different fields in order to generate an evidence base for companies that seek to improve their employee’s health. Medical doctors from sports medicine and cardiology, GI scientists, planners, traffic engineers and mobility consultants collaborate in a highly inter-disciplinary setting. The research project is funded by the Austrian ministry for transport, innovation and technology in the program MdZ .
The project’s main idea is the following: commuting to work is time-consuming and if done by private car bad for many reasons: congestions, noise, air pollution, space consuming, expensive and inactive. The project aims to tackle the last aspect and provides highly detailed information for companies about which return they can expect from investing into employee’s active mobility. It is important for employers to get an idea how effective different interventions are. On the other hand, employees can only be motivated to change well established commuting routines when the alternatives are realistic and attractive.
This is why we (a) started a clinical intervention study, (b) developed advanced routing algorithms and spatial models and (c) pack all the information into an intuitive, interactive information platform.
Lots of activities have been going on during the first twelve months:
A clinical intervention study with 70 subjects was designed. The study was approved by the responsible ethic board, before it was implemented in a large company.
The study design is the following: 70 car commuters are randomized either into two intervention groups or into a control group. In one intervention group subjects are motivated to switch to bicycle commuting. Subjects in the other intervention group switch to public transit and walking. All subjects are medically investigated before and after the intervention. Additionally, all subjects are required to document their commuting mobility in a diary. In order to validate this documentation and to derive estimations for the energy turnover, the subjects wear GPS fitness watches for two weeks in the beginning and another two weeks towards the end of the intervention.
The aim of this study is to estimate the health effect of active mobility interventions, which can be implemented in any company.
In order to recommend realistic routes for active commuting, we developed a sophisticated routing workflow, which makes use of a national, multi-modal routing service (VAO ). The routes are optimized in terms of health (minimum distance for walking or bicycling) and travel time.
Together with the routing recommendations users of the platform are provided with spatial information about the quality of the environment. For this, we have developed spatial models that calculate walkability, bikeability and PT quality indices and map them at a very high spatial resolution.
Currently, the project partner TraffiCon is developing the concept for the web-based information platform. A first proof-of-concept will be presented at next year’s TRA conference in Vienna. The platform will provide detailed information on health effects of active commuting, recommendations for individually optimized routes and information on potential interventions for companies.
Two master students at Z_GIS have started to analyse the GPS and heart rate data from the first data collection phase. First results look very promising with regard to mode detection and trip parameters.
The clinical study is going to run until May 2018 and first results are expected to be available soon afterwards.
As project leader I’m happy to say that the collaboration with partner from very different domains is extremely fruitful. Actually, we learn a lot from each other and it becam obvious that there are a lot more common interests (“What have GI scientists have to do with cardiologists?!”) than we had expected!
In order to share our experiences and to learn from others who are doing research in similar settings, we will organize a special session at the GI-Forum conference 2018 in Salzburg. The call for papers has already opened – you should definitely have a look at it.
A recent publication of the national klimaaktiv program caught my attention last week. Following this report, 55% of all kids in Austria are brought to kindergarten by car. The share of kids cycling to kindergarten is below 5%.
As the figures are for Austria, I don’t know if they are representative of my home town. Unfortunately, I haven’t found any statistics for Salzburg. My personal experience is that the number of kids on bicycles is very low – at least in my kids’ kindergarten. Most of the time we are the only one coming by bike.
Independently from the availability of statistics, open (government) data allowed me to do a bit of spatial analysis on the accessibility of kindergartens in Salzburg. The results are striking: there is virtually no need to bring kids to kindergarten by car in terms of distance or travel time. However, the environment of kindergartens is not always pedestrian- and bicyclist-friendly and of course, this impacts the mode choice for bringing and picking-up kids.
For the analysis I used the following data and settings:
- Address data are available from the federal Bundesamt für Eich- und Vermessungswesen. The dataset can be downloaded from their website .
- The location of kindergartens are published as OGD by the city administration. The most direct way to the data is to use this WFS . I need to add a disclaimer here: I’m not sure whether all kindergartens are included in this data set. Probably, not all private facilities are on the list.
- For the network analysis I used authoritative road data (GIP ), published as OGD and enriched this dataset with additional information, which is also available via the national OGD portal .
- I used the routing engine from ArcGIS Desktop 10.4 for performing the network analysis.
For the optimization of bicycle routes I employed the impedance model we have developed a few years ago for a bicycle routing service (see here and here for details). An average speed of 15 km/h (which might be a little bit too high) and global turn impedance at intersections were used as input parameters. For the pedestrian routes the shortest path was calculated and an average walking speed of 0.8 m/s was assumed.
- The walkability and bikeability index were modelled in the context of the GISMO research project; details are going to be published soon.
The average walking time to the next kindergarten, considering all address points in Salzburg is 8.14 minutes with a standard deviation of 6.10 minutes. The median is 6.86 minutes. 74% of all address points (16,741 of 22,694) are within 10 walking minutes from the next kindergarten. This means that 3 out of 4 kids could walk to kindergarten within a reasonable time.
These figures are even more striking when kids cycle to kindergarten. The average travel time is 2.90 minutes (!) with a standard deviation of 2.21 minutes. The median is 2.42 minutes. From 89% of all address points in Salzburg the next kindergarten could be reached in less than 5 minutes. This figures rises to incredible 98% for a maximum travel time of 10 minutes. Plotted on a map these figures look like this:
Although these numbers are striking, obviously, there are good reasons for parents not to let their kids walk or bicycle. Interestingly, the scientific evidence on influential factors on the mode choice for kindergarten-related trips is weak (not to say nonexistent). Studies of school pupils’ commuting trips reveal a significant impact of parent’s perception of the suitability of the environment on the mode choice (see for example Timperio et al. (2004) or Bringolf-Isler et al. (2008) ). The impact of parents’ perception and behavior might be even bigger for kindergarten kids. In any case, the quality of the environment needs to be high in terms of pedestrian- and bicyclist-safety if active mobility should be further promoted. That this is currently not always the case, becomes obvious in the following maps:
Last year, results of a large study on the relation of urban environments and level of physical activity were published in the prestigious medical journal The Lancet . The authors found that approximately 50% of the WHO recommendation of 150 minutes physical activity per week can be stimulated by an adequate environment.
Thus, the provision of safe and comfortable infrastructure has a huge effect, far beyond kids walking or cycling to kindergarten. Usually, this effect is quantified in terms of health, environment or economy. But apart from these important dimensions there is another one when it comes to kids. As Moran et al. (2017) proved earlier this year, the mode choice influences navigation skills and (spatial) knowledge of the neighborhood. As a geographer I’m tempted to conclude that these findings are a sufficient argument for further promoting active commuting to kindergarten and school!
Today, I had the honour to chair another special session that dealt with GIS and mobility research at this year’s GI-Forum conference . The session “Spatial Perspectives on Active Mobility” was the third in a series (see here for a review of the 2016 and here for the 2015 session).
Of course, we will have a “Spatial Perspectives on …” session in 2018 again – the call will be published in December this year. So, consider this as an option for your publishing and dissemination strategy (by the way, the GI-Forum journal is open access!)
This year’s special session was a paper session with four speakers, who all went through a rigorous review process. The diversity of the contributions was high, demonstrating the wide range of mobility research where GIS plays a crucial role:
- Irene Fellner from Vienna University of Economics and Business opened the session at the very local scale. She presented her work on landmark-based indoor navigation. Although the applied ILNM (“indoor landmark navigation model”), an extended version of Duckham’s et al. (2010 ) LNM, performed well, Irene pointed to two major challenges: first of all, the ILNM requires very detailed data, which are not always available and secondly, the visibility of landmarks from the perspective of the user is not always given or unknown.
Irene’s paper emerged from her master thesis at the University of Salzburg, where she successfully finished the UNIGIS MSc study program. Dr. Gudrun Wallentin, UNIGIS program director, regarded this special session as perfect stage to hand over the UNIGIS International Association (UIA ) award for excellent master theses. Congratulations!
- Ulrich Leth (Vienna University of Technology) presented the findings of a recent study where they investigated the impact of a bike sharing system on public transit ridership in the city of Vienna, which is famous for its extensive and well-performing public transit system. In total, Ulrich and colleagues analysed 1 million Citybike trips from 2015. Different to the expectation the title provoked, they found that the bike sharing system virtually has no impact on PT ridership, simply because of the huge difference in size and capacity. However, some details in their results were interesting and probably of relevance for other BSS: a) Citybike trips primarily substitute short and inconvenient PT trips, b) most bike sharing trips are made when the travel time ratio compared to public transit is 0,5 and c) the most popular OD relations are typical student trips (between transport hubs and university and student dormitories and transport hubs or universities).
- Tabea Fian, a student from Georg Hauger’s (lead author of the paper) group, also from Vienna University of Technology, presented a spatial analysis of urban bicycle crashes in Vienna. Interestingly, the data were very similar to those I’ve extensively used in my PhD (see this paper ). In a purely exploratory study design Georg has tried to identify blackspots in the network and tested for their significance. However, as it became evident in the discussion, final conclusions are hard to draw without a statistical population.
- The last presentation was given by Anna Butzhammer from RSA iSpace. She presented parts of her excellent master thesis, in which she analysed the inter-modal accessibility of central places. For this, she developed a model that facilitates door-to-door travel time calculations with different modes. Her findings are especially important for planning and optimizing public transit systems, which can be regarded as backbone for sustainable mobility.
Tomorrow, the German-speaking sister conference, AGIT , will host a special forum on autonomous driving and on Friday I will chair another session on advances in GIS-T. Well, there will be a lot to discover, learn and discuss; if you don’t have the chance to be there physically, follow me on Twitter and stay updated.
Our colleagues from Salzburg Research (SR) are very active in the field of floating car data generation, management and analysis. Among others, this real-time traffic status service is fed by their data.
In order to establish a community of researchers, authorities and companies around the topic of floating car data, SR hosts the annual “FCD Forum” in Salzburg. This year, I had the honor to contribute to the program . Since we have been working a lot with bicycling data over the last years, I was asked to evaluate the potentials of a conceptual transfer from FCD to “Floating Bicycle Data”. Well, a very fundamental finding in my research is that the term “Floating Bicycle Data” is not established yet in the scientific literature. Thus, the term is to be regarded as a word game derived from the forum’s agenda. However, I think it makes perfectly sense to invest some efforts in this context.
In my presentation , I started my argumentation from the fact that a) bicycle traffic is a relevant element of urban mobility, b) the modal share is likely to increase in the next years and c) a sound evidence base is required for future investments in bicycling infrastructure.
Currently, very little is known about the spatial and temporal distribution of bicycle traffic within cities. Comparably few permanent counting stations, sporadic, punctual counting campaigns and irregular mobility surveys do not provide sufficient and reliable data to support evidence-based policies on the local scale level. On the other hand, the popularization of the “humans as sensors” concept (Goodchild 2007 ) has opened new possibilities to acquire data on bicyclists’ movements in urban networks. When talking about floating bicycle data, I used it as a catchy term, which summarizes all kind of geo-located movement data from bicyclists; they don’t need to be necessarily in real-time.
As I’ve shown in my presentation, there a numerous application examples where floating bicycle data would make perfectly sense. However, there are several conceptual challenges, which need to be considered (most of them are also relevant for floating car data):
- When floating bicycle data are harvested through crowd-sourcing applications the data are not necessarily representative for the entire population. I referred to participation inequality or the 90-9-1 rule (see Nielsen 2006 ) in this context. Additionally, different apps are used for different purposes. Thus, the data might be biased for example towards leisure trips (as it is the case with Strava data in Salzburg).
- Currently, there is no common data standard and the heterogeneity of bicycle mobility data is huge. Good news in this context were published earlier in this year by the European Commission (see this report from the COWI project).
- Since there is no obligation to register bicycles, the (spatial distribution of the) total population is unknown. Consequently, it is hard to estimate the total bicycle traffic volume from samples. In contrast to that, cars are registered and at least the car holders’ address is known.
- In order to further process movement data (GPS trajectories), a sound and very detailed reference graph is required for map matching. In most cases network graphs are not available at this level of detail (this holds true for authoritative data as well as for OSM). Consequently, GPS trajectories can only be matched to center lines at the moment.
Although this selection of challenges might be regarded as obstacle for a broader engagement (I prefer to interpret them as research opportunities), I expect the topic of floating bicycle data to emerge in the coming years for a simple reason: the market for floating bicycle data is definitely smaller than for floating car data. But, bicycle traffic is already a major element in urban traffic and its share will become even more substantial in the next years. As a consequence, cities need to invest in adequate infrastructure and these investments will hardly be made without a sound evidence base. Floating bicycle data could close a significant gap in this regard.
If you are already working with floating bicycle data (but haven’t used the term yet), have ideas on how to further push the topic or simply want to comment on the concept, please do not hesitate to contact me! I’m happy to learn from your expertise.
For those who are about to write a thesis in this or a related context, have a look at this proposal .
After several months of setting the stage and doing lots of preparatory work, we are currently entering the ‘core phase’ in two research projects at the GI Mobility Lab . In this context we provide the opportunity to Master’s students to participate in the projects and write their thesis in GIScience (or related fields).
Our part in the FamoS project is, among others, to develop an agent-based bicycle flow model for an entire city. In this context we offer two topics:
- Behavior to space (description )
- Exploring geoprocessing, geovisual analytical and mapping functionalities of GAMA (description )
Experts from sports medicine, GIScience and transport planning and management are collaborating in the GISMO research project in order to provide a sound evidence basis for the promotion of active commuting. Part of the research is a clinical study, in which we document the subject’s mobility by different means. For the analysis of this data we offer the following two topics:
- Analysis of movement data from fitness watches (description )
- Linking travel diaries and GPS trajectories (description )
Originally, this blog was intended to document the progress of my PhD research. Mhm, this goal has been successfully reached yesterday …
I finished my doctoral studies with a thesis on Spatial Information and Bicycling Safety and yesterday’s defense. The thesis is based on five peer-reviewed, published papers and aims to strengthen the spatial perspective in bicycling safety research.
The thesis is motivated by the fact that bicycling safety research is dominated by non-spatial domain experts, e.g. with backgrounds in trauma medicine, psychology, bio-mechanics, sociology, epidemiology, engineering, planning, law and some more. Interestingly, the spatial perspective on bicycling safety is hardly ever considered in these domain-specific approaches. This holds especially true for bicycle crash analyses, where basic geographical concepts, such as nearness, spatial autocorrelation and topology, are hardly ever considered.
Neglecting location as a co-determining attribute of safety is remarkable for a very simple reason. Mobility of people – and thus bicycling – as such is spatial by its very nature. Consequently, bicycling safety (from the physical environment to crashes to individually experienced safety threats) has spatial facets, which can be modeled and analyzed accordingly in order to gain relevant information for safer bicycling.
The primary hypothesis of my doctoral thesis is that spatial models and analyses contribute to a better understanding of certain aspects of bicycling safety and provide relevant results, which support measures to mitigate safety risks for bicyclists. Specifically I argued that:
- Geographical Information Systems (GIS) facilitate holistic approaches for improving the bicycling safety situation. The spatial perspective is relevant for virtually all stages of the implementation of bicycling safety strategies.
- Model-based approaches have a great potential in safety assessment and can form the basis for a number of applications – from status-quo analysis to planning and route optimization.
- The spatial analysis of bicycle crashes reveals significant and safety-relevant patterns and particularities, which remain hidden in common, non-spatial or highly aggregated approaches.
- The spatial perspective is crucial for advanced (simulation) models, which are necessary for reliable risk estimations on the local scale. Furthermore, the spatial implications of risk mapping on the local scale must be made explicit.
The thesis is structured in three elements. The first paper demonstrates the contribution of GIScience to bicycling safety research and is intended to set the stage for the remaining papers. Two of them primarily deal with spatial models in the context of road space assessment and transport modeling, while the rest is about spatial analysis of bicycle crashes.
Although the completion of my doctoral studies is a huge, personal milestone, there is still a lot of research work in this context to be done. Besides the further development of the spatial models, the applied statistical methods and analysis routines, I see research gaps in the context of data (from static to dynamic real-time data and data streams), information (e.g. what are the effects of information provision on decision process or on individual and collective behavior?) and cross-domain collaboration.
The amount of work that still lies ahead motivates me to further blog on some of our research activities and to connect with anyone who is interested in spatial information, bicycling safety, urban mobility etc. I’m looking forward to learning, reading and hearing from you in virtual and – even more preferably – in face-to-face communication!