Abstract
To encourage cycling for both recreation and transport, Singapore continues to plan and develop cycling path network infrastructure. Recent developments in the use of big data have shown the potential of crowdsourced data and geolocation-based services in cycling planning. One source of data is Strava, which is a popular fitness and social network app for athletes, where users record activities including cycling trips using their smartphone and share these trips publicly online. The study aims to collect cycling data from Strava and create a user-defined cycling route network. Cycling data from individual trips was collected from the Strava website using a web scraping Python script and analyzed using ArcGIS. Data from 3,653 cycling trips was obtained, including trip data and GPS route tracks. Most of the trips were determined to be sport cycling activities, characterized by fast speeds and long ride distances which are mostly home-based roundtrips. Raster analysis was able to visualize utilization of cycling infrastructure and street networks across Singapore, highlighting several intertown cycling routes across Singapore which are popular for recreational and sport cycling. New, external routes outside of the existing cycling infrastructure and street network revealed trails and routes which are popular with cyclists. Origin and destination data was also analyzed from the GPS data, which reveals a spatial mismatch between OD points and present access to cycling infrastructure. Overall, the study was able to use web scraping to obtain Strava data which provides many applications and opportunities for cycling planning not just in Singapore, but in cities worldwide.