Tagged: safety

Bicycle routing and safety

Back in 2016, I’ve presented a review study on bicycle routing portals and safety as routing criteria at the International Cycling Safety Congress in Bologna (ICSC internet).
Now, 16 months later, I’m happy to announce that an extended version of this contribution has been published in the Open Access journal Safety (follow this link internet to read the paper). Together with Hartwig Hochmair internet from the University of Florida, we updated and extended the study and derived recommendations on how to better address prevalent safety concerns in trip planners.

Safety concerns are still a major barrier for a larger bicycle mode share in everyday mobility (see e.g. Wegman et al. 2012 internet). In this article, we hypothesized that current online trip planners do not sufficiently consider safety as decisive criteria for route choice. Given the fact that  many cities around the world are still lacking of adequate, connected infrastructure, we consider trip planners as a potential element to bridge the gap between safety concerns and the built environment. Routing optimization could be used to highlight and recommend safe routes.
However, we have learned in our review study that virtually none of the investigated portals explicitly provides safety as a routing criteria. On the one hand this might be due to liability reasons, but on the other hand there are a few examples, which optimize routes in terms of safety (see this example internet, where the recommended route is explained as safest connection).

Factors that contribute to the routing criterion “safety” according to Hochmair 2005 internet (orange), and general routing criteria (green). Both classes are mapped to the analyzed portals in the matrix (right).


Lecture series “Active Mobility”

Since the VeloCity internet conference took place in Vienna in 2013, the Institute of Transportation internet (Vienna University of Technology) hosts an annual lecture series on bicycling and active mobility in general.

This semester, 80-100 students from various planning domains (urban, transport, regional planning) are attending the weekly lecture on “Active Mobility” internet. Yesterday I had the privilege to present parts of my current research and provide an overview of potential contributions of spatial information to an enhanced bicycling safety situation (slides in German language):

Although some of the students have already worked with GIS, none of them employe GIS in the context of mobility or transport research (at least nobody raised his/her hand when I was asking). Thus, I was happy to serve an appetizer for introducing the spatial perspective to a rather “technical” planning community.

VeloCittà bikesharing & POLIS conference

We contribute spatial information to the design and optimization of a city-wide BBS.

We contribute spatial information to the design and optimization of a city-wide BBS.

150 participants from 23 countries gathered on November 30th in Rotterdam to attend the VeloCittà internet bikesharing conference, which was held in conjunction with the annual POLIS internet conference (450 participants, according to the organizers). While the VeloCittà conference was exclusively dedicated to bikesharing, the POLIS conference offered a broader perspective on sustainable transport. I was in Rotterdam primarily for the POLIS conference because I had a presentation, but it was also a great opportunity to get an impressive update of recent bikesharing practice and research. Lot’s of what I’ve learned can be directly linked to our current involvement in the planning of a bikesharing system in Salzburg, Austria.
All presentations of both conferences can be found on the respective websites. Thus, I will focus only on two topics I’ve found especially relevant for our research and project work.


Willemijn Lambert (@WM_Lamber internet) captured the essence of the VeloCittà bikesharing conference.

Success factors for bikesharing systems

polis2016aIn a very interesting session at the POLIS conference on sharing systems, Sebastian Schlebusch from Nextbike internet gave some insights into the company’s history. Several years they were treated quite harshly by public transit operators who feared for their business. However the break through of bikesharing systems (BSS) came. In accordance with Sebastian’s talk the following success factors occurred in various presentations at both conferences:

Cologne's bike sharing system (KVB Rad) is integrated in the city's public transit service.

Cologne’s bikesharing system (KVB Rad) is integrated in the city’s public transit service.

  • Political support. Obviously this seems to be the most decisive factor for successful BSSs in any country.
  • Integrated systems. An increasing number of cities regard bikesharing systems as an element of public transit services. This is reflected in the planning of the network, pricing and promotion. Cologne’s BSS internet is a good example for a large, integrated system.
  • Robust business models. This factor becomes important when initial subsidies fade out. Alberto Castro internet, one of the keynote speakers at VeloCittà, demonstrated how fast BSSs without sound financial (and operational) basis disappear internet.
  • Appropriate planning. Nicole Freedman, keynote speaker at VeloCittà, made a compelling case for the importance of realistic projections and tailored BSS design. Cities are comparable only to a certain degree and thus, BSSs cannot be simply transferred. Specific (mobility) characteristics of cities, from PT service level to topography, need to be taken into account.
  • User-tailored, easy solutions. The needs and expectations of users must be addressed in every aspect: from intuitive interfaces for initial registration to the ease of handling the hardware.
    To know and consider people’s reasons for not using BSSs is especially valueable when systems should be improved. In many cases the barriers for BSS usage can be lowered or removed with small adaptions.
  • Visibility in public space. In order to raise awareness for bikesharing it is necessary that the system is visible in public space. This visibility can be achieved by an appropriate station design, but also with art in public space.
  • Make it beautiful. Directly associated to the latter point Nicole Freedman strongly argued for aesthetically pleasing, beautiful bikes and infrastructure. Way too often BSSs are shaped by technicians and technology. With a good design of hard- and software people can be made curious; once they are attracted to the system, the possibility is high for turning prospective into active users.

polis2016cAt both conferences lots of case studies were presented. At least two of them were really remarkable:
Krakow (~ 760,000 inhabitants) initially launched a system with 30 stations and 300 bikes, which turned out to be not that successful. Thus, the city relaunched the entire system under a new name (Wavelo internet) and with 1,500 bikes at 150 stations, which is above the average bikes per people ratio in Europe (ref. OBIS internet handbook)!
A much smaller, but very successful BSS can be found in Pisa (CICLOPI internet). Marco Bertini presented the city’s strategy to make people in Pisa love their bikesharing system: “Bikesharing is note a service for citizens, but part of the community.” With this approach Pisa achieved impressive key figures: 5-8 rides per bike and day, virtually no vandalism and not a single bike stolen in 4 years.

Road Safety

polis2016dMore people are killed in road crashes than by malaria or tuberculosis, according to a recent OECD report internet that calls for a paradigm shift in road safety. Before this background and with a special focus on the role of large cities the International Transport Forum (ITF internet) launched the Safer City Streets internet project, which was presented by Alexandre Santacreu. The aim of this project is to provide an environment for exchange of data, experience and knowledge. What I regard as an asset of this project is the drive to publish data as OGD.
Alex pointed to the difficulty of comparing data from various sources, especially when crashes of vulnerable road users are investigated (different reporting procedures, classification, under-reporting etc.). Of course, this is nothing new, but my impression is that the limited comparability of data is mostly neglected in analyses of global data (I’ve demonstrated an aspect of this in this post internet).

While the Safer City Streets project operates on the global scale, the Netherlands have launched a national project where cities can learn from each other with respect to crash prevention and safety measures. Charlotte Bax from SWOV internet presented this benchmarking project that is built upon the three elements comparing – learning – improving. Two aspects caught my attention: (1) None of the data are made public because the involved city administrations fear the pressure that might be put on them after publishing crash details. (2) Even in the Netherlands’ city administrations struggle to make use of their data; Charlotte referred to cases where responsible departments were not able to tell how many kilometers of bicycle infrastructure they had.

Benchmarking on the very local level was at the core of Eric de Kievit’s internet presentation on the development of a compound road safety assessment. For this, two approaches were combined. Firstly, a network safety index, which consists of an enormously detailed description of the road space (every 25 meters the road profile was investigated based on street view photos). And secondly, a safety performance indicator that focuses on road user’s behavior. Both perspectives are then used as basis for targeted infrastructure measures, law enforcement, education and communication campaigns.

My own contribution to the session on road safety was about spatial analysis of bicycle crashes on the local scale level. The presentation was a synthesis of two of my latest journal papers (JTRG internet and Safety internet):


In both conferences it became evident that there are lots of innovative and creative solutions for promoting sustainable mobility in urban environments. However, there is no philosopher’s stone that solves all problems immediately, but cities all over Europe have to work hard to make progress.
I have the strong impression that the discussion and collaboration across domains and institutions is a key for sustainable solutions for cities. Urban environments are complex and thus require multifaceted strategies. In any way, we are ready to contribute spatial expertise for the good of our cities and their citizens.

GIS and bicycle safety

Today and tomorrow the “4th International Cycling Safety Congress” (ICSC internet) takes place in Hannover, Germany. This is an excellent opportunity to learn from and connect to experts from various domains. The common denominator is, as the conference name indicates, bicycle safety, but the approaches presented so far are rather diverse.
Today’s presentations ranged from medicine (which body parts are most often injured in different types of crashes), to legal aspects (single- or bidirectional cycle paths), to hardcore technology (automatic analysis algorithms for videos from naturalistic studies) to planning (optimal road design, especially at intersections) and to social and psychological aspects (by which values and attitudes are ‘cyclists’ driven).

What was largely missing so far was the spatial perspective on bicycle safety. Hence I’m looking forward to add a nice, little piece to the multi-disciplinary mosaic tomorrow.
My argumentation starts from the fact, that bicycle accidents are spatial by their very nature. They don’t occur in the nowhere, but in geographical space. Thus they can be analyzed in Geographical Information Systems (GIS). Through geospatial mapping and analysis geospatial patterns and dynamics become obvious, as the figure below demonstrates.

Spatial dynamics of bicycle accidents in the city of Salzburg between 2002 and 2011.

Spatial dynamics of bicycle accidents in the city of Salzburg between 2002 and 2011.

From bicycle crash frequency to risk estimation.

From bicycle crash frequency to risk estimation.

Although the spatial patterns and dynamics that emerge from simply putting the crash locations on a map are interesting and relevant for multiple application contexts, the estimation of risk is even more helpful. The problem here is, that exposure variables are hardly ever available beyond the city scale.
In order to account for the variation of crash occurrences and the corresponding risk within the city we made use of the recently developed (and published internet) bicycle flow model for the city region of Salzburg. Relating the city’s reported crash locations to the simulated bicycle traffic volume results in an estimation of risk at the smallest possible scale.

Besides the analysis of (historical) crash data, Geographical Information Systems are also capable to model and simulate potential safety threats. For my presentation I’ve selected three use cases in which such a geospatial model (Loidl & Zagel, 2014 internet) is employed in decision support and planning tools:

  • The first use case is about the quality of accessibility of central facilities, such as university buildings. The introduced assessment model allows for an evaluation of the immediate environment of the respective facility. Based on this status-quo analysis, existing corridors and barriers as well as potential connections can be identified.
  • In the second use case, the assessment model is fed into an interactive, simulation environment (see Wendel, 2015 internet for further details). Through a web interface the effect of infrastructural or legal measures on the overall safety index can be tested very intuitively.
  • As a third prototypical field of application, besides status-quo analysis and simulation, the assessment model is employed in a bicycle routing service. Operated by the city and the federal state of Salzburg, the web portal radlkarte.info internet recommends the optimal (safest, most comfortable) route for utilitarian bicyclists.

What all examples, shown in the presentation, have in common is that they make use of the spatial characteristics of bicycle crashes and the transport network respectively. Geographical Information Systems can thus be considered as well-performing integration platform for multiple perspectives on the complex system of bicycle safety and as facilitator for innovative planning and information tools.
Have you got any comments, ideas or critical remarks? Feel free to write a comment, use the contact form or simply get connected on Twitter internet.

Selling snow to eskimos

… or talking about urban cycling in Oldenburg.

Today I participated in a workshop on sustainable mobility which was held in the context of this year’s EnviroInfo internet conference. There I had the chance to present (slides here internet) our research on bicycle safety as fundament of a comprihensive bicycle promotion strategy.
Before I came to Oldenburg I was aware that it has an amazing bicycle modal split, but what I saw was much more than I had expected. In Oldenburg it’s not about painting a cycle lane here or limiting traffic speed there. All cycle ways are separated from the main road and worthy of their name. What’s the secret of Oldenburg’s success? Maybe this picture says more than thousand words:


Are bad public transport connections (look at the schedule!) and an excellent bicycle infrastructure the key to the amazing bicycle modal split?

Another thing that astonished me was a statement by Dr. Hermsmeier from EWE (utility company) who said in his morning keynote that EWE collects tons of data with their smart meters but don’t know what to do with it! Can you imagine this?! I always thought that the only reason for installing smart meters was to provide intelligent, user-tailored services (official version) and to generate detailed, individual profiles of all consumers (business model). But now after hearing this from the head of the R&D department it seems to be simply a waste of money – at least as long as the smart meter is not from Nest internet.

BTW, the keynote in the afternoon by Chris Preist (University of Bristol) was my personal highlight at the conference. His big picture of the manifold impacts of ICT on the environment was really inspiring. I’ll need to browse is publication list internet for more details … The main point he made was, that ICT accelerates societies and that it is of crucial importance to be clear about the target. My personal target now is to catch my train back home!

Lot’s of activity behind – more ahead

As this year inevitably comes to an end, it’s time to summarize my personal bicycle-related highlights of 2013 and try to sketch what lies ahead.

First things first: I’ve officially started my PhD project last summer and I’m very thankful to those who make this next step possible. During the last few years I’ve loosely worked on several mobility-related topics. Primarily on bicycle routing applications. Now the time has come to bundle up all the findings and developments and to focus on concrete research questions.
My PhD project has the working title “Analyzing, modeling and assessing bicycle road safety with GIS” and can be split up into three sub-topics:

  • Systematic analysis of bicycle accidents in an explicitly spatial context.
    For this I use bicycle accident data for the years 2002-2011. The 3,096 incidents in the city of Salzburg/Austria are georeferenced and contain all details reported by the police. The biggest challenge in this context is the absence of any statistical population. Hence it won’t be possible to calculate any risk exposure. But still, for anomaly detection the data seem to be very promising.
  • Development of an road network assessment model.
    Based on previous research, I’m going to refine the indicator-based assessment model (Loidl & Zagel 2013 internet, in German). This approach makes use of the power of geographical information systems and allows for a global assessment of road network quality.
  • Application of the assessment model in a planning context.
    The indicator-based assessment model should be adapted in a way that it can be employed in a traffic planning context. The idea is to use it for weak-point analysis and for simulation purposes.

The research conducted for this PhD project is widely triggered by funded projects. Thus the ideas as such are rooted in real-world problems. In turn, all generated results can be tested and applied immediately in practical environments. Just like this project …

radlkarte.fwA very nice application which was launched in 2012 has been extended this year. The web-based bicycle routing application “radlkarte.eu” is now available for the city of Salzburg and its neighboring municipalities in Bavaria. The application primarily addresses commuters who use their bicycles as an efficient, flexible and healthy mobility alternative. Funded by the city administration of Salzburg, the federal state of Salzburg and the EU Interreg IV program, the routing application is an integral part of the regional long-term bicycle policy.
 Z_GIS’ contribution to the application comprises the data management and the provision of the routing network (the web services has been developed by a partner company, TraffiCon internet). The innovative part of the routing application is the explicit consideration of bicycle safety as a routing criterion. For the calculation of this safety index the indicator-based assessment model has been used. Here the challenge was to apply the model for different data sources and data models. Above that, the model had to be adapted for urban and rural environments. With several lines of code we managed to process the whole workflow automatically. This made it possible to run fast calibration cycles.
The result of this project is a seamless, cross-border routing application for bicyclists who are not only interested in the shortest but also in the safest route.

As my department, the department of geoinformatics, Z_GIS internet, has been established as an independent department just recently, this year has been a time of consolidation. In this context I’m very happy that one of the eight, newly defined working groups has been dedicated to mobility and transportation research. In this institutional context research on bicycle safety will play a significant role. This directly brings me to a brief outlook …

At the moment we are waiting for the definite start of a project which will allow us to further improve the bicycle routing application “radlkarte.eu”. The national funding agency (Klima- und Energiefond) has already allocated funds, but one of the two institutional sponsors still needs to manage some internal ambiguities. Anyway, we are hopeful to kick-off the project in early spring.
Beside this we are currently writing on research proposals and – of course! – hope to be successful!

I’m planning to publish a first journal paper about the accident analysis in the first half of the year. Together with two brilliant colleagues I need to finish the spatio-temporal analysis and wrap up the methodological context. Depending on how much project work is to be done, I’d like to start with smaller papers about some aspects of our last projects (e.g. harmonization of different data models for routing purpose).
The indicator-based assessment model is planned to be reconfigured. At the moment the result of the assessment model is static. This means for example for the routing application that user inputs cannot be considered; the safety-index is preprocessed. In a next version the assessment model should allow for user inputs and do the processing during runtime. In this context the standardization of the index values and the balancing of the indicator weights will be the major challenges.

During my first (half) year as PhD student I tried to figure out how I can manage my daily work for UNIGIS, all the project work and my own research/publishing. In some respects I definitely need to improve my routines! But I hope that the wheels – not only of my bicycle – keep turning and significant progress can be made in 2014.