Assessing the Potential of Big Mobility Data to Improve Transport Modelling for Cyclists
Georgios Christou, Jacobs Engineering UK
The paper is associated with the investigation of the potential use of data acquired from activity tracking mobile applications for the purposes of better transport modelling for cyclists.
Cycling is a healthy activity that helps cities around the world to reduce their congestion without any environmental impact. Despite the fact that cycling has been promoted among cities of various sizes, several problems associated with the safety of roads may concern potential bicycle riders.
This study is associated with the investigation of the potential use of data acquired from activity tracking mobile applications, such as Strava, for the purposes of better cycling transport modelling. Data, licensed from Strava, were acquired and consisted of 1,800,000 rows (time, origins and destinations) which were used for the creation of multi-temporal Origin – Destination (OD) matrices for cyclists in London and the results were compared with Census travel to work statistics. The created OD matrices describe the commuting movements of cyclists for the year 2013, for each month and for three time periods during the day (AM-IP-PM). Strava trip ends were compared against census travel to work statistics, observing similar patterns. The creation of multi-temporal matrices showed variance of bicycle usage between different months of the year and linear correlation with the temperature. The paper challenges the usage of survey data of transport modelling for cyclists, showing that the use of survey data alone may result in an unrepresentative impression of cycle demand. Furthermore, a factor was calculated for each month of the year, proposing that survey data collected during different times of the year, should be multiplied by those factors, in order to estimate the maximum demand of cycling. Alongside this, data sets like Strava, can give a dynamic view of the city, allowing sustainable design. The study recommends further research into this kind of data sets for the potential use of crowdsourced data. This could lead to the creation of improved traffic simulation models for cyclists, and aid to the design of an improved, safer and sustainable road network for cyclists.
Keywords: urban mobility, cycling, crowdsource data, sustainability, transport modelling, big data
Association for European Transport