Evaluation methodology on comparable data collection on walking and cycling.

Scope of the methodology 

The study addresses the lack of comparable statistics concerning walking and cycling in the EU. Travel surveys are the main source of statistics on average daily walking and cycling distance and number of trips per person. Countries were divided into seven groups according to the data collection and reporting methods used. The most recent and representative statistics were used to produce a first comparative overview on walking and cycling use based on these non-harmonised statistics. The European median daily walking and cycling distances and number of trips were estimated. The results were compared with statistics derived from EU surveys. The potential of using crowdsourcing for data collection of infrastructure was examined and results were promising for collecting cycling infrastructure statistics. This would require standard definitions and guidelines for the data collection. The potential of big data was analysed through the case studies of ‘Google Better Cities’ and ‘COWI City Sense’. Although promising for the future, neither of these could provide active modes statistics. Recommendations to further improve information and statistics on active modes include: common definitions, new indicators concerning walking friendliness of urban environments, alignment of data collection efforts with ongoing and future initiatives and post-processing methods to improve comparability. 

Key methods

This study began with active mobility data mapping across European cities and states to understand how to improve harmonization in data collection and processing. It provides some general comparisons of the average daily distance walked, number of walking trips per person, comparison of the average daily distance cycled, number of cycling trips per person, and total distance walked and cycled per year in the EU.  

The study also grouped countries into seven main categories based on which countries carry out regular surveys, whether or not countries report key indicators, and whether surveys were only for the capital region or for the entire country, and if travel surveys included actives modes. The report details the many issues of collecting active mode data and provides recommendations for various stakeholders. It provides information on how to collect data about quality of infrastructure. It also discusses the potential of big data for collecting walking and cycling data: Both ‘Google Better Cities’ and ‘COWI City Sense – Signal Re-identification’ appear to be promising for data collection of active modes use in the future. 

Demonstration results 

Recommendations on possible data collection strategies:

• Common definitions should be developed starting at a very basic level:  “what is a pedestrian?” and “what is a cyclist?” There is currently a lack of consistent or explicit definitions in many active modes data collections at country and city level. 
• The definitions used for sampling and data collection must be explicit, particularly regarding the term ‘urban population.’ The questionnaire responses revealed varied interpretations of urban population size, highlighting the lack of a clear definition across countries.  
• The definition of a trip is another key issue. Essential questions on how trips are recorded (e.g., main mode vs. all modes) and the boundaries of trips (e.g., core city area vs. greater city area) are crucial. 
• Collect walking and cycling pkm/day and trips/day at city level in Europe making use of existing initiatives: A possible synergy with existing initiatives could be the “Quality of Life in cities” survey conducted every three years since 2004. 

Recommendations on data management and analysis: 

• Develop post-processing methods for comparable statistics: When harmonised transport statistics are available, the potential to use them for comparison or interpretation of active mode statistics collected by countries and cities should be examined.  
• Seasonal variations and climate conditions should also be considered: optimal cycling conditions are not the same in Scandinavia and the Mediterranean countries.  

Stimulate and harmonise crowd sourcing active modes data collection: To improve the mapping of cycling infrastructure in Europe, crowd-sourcing initiatives like OpenStreetMap (OSM) should evolve their mapping rules to align with the improved definitions. Additionally, participation in mapping cycling infrastructure should be stimulated, potentially through successful initiatives such as the European Cycling Challenge.  

Follow developments in big data collection: Improvements to further resolve bias issues should be followed up in Google Better Cities. Such a method could also be combined with satisfaction surveys.