The aim of this project is to analyse patterns of usage in dock-based bike sharing systems in order to distinguish between different types of stations with the end goal of being able to predict the average daily traffic patterns of stations based on spatial factors in their service areas. The analysis is based on trip data from bike sharing systems in New York City, Chicago, Washington DC, Boston, London, Helsinki, Oslo, and Madrid as well as other data external to the systems.

In the analysis, different clustering algorithms are introduced to cluster stations based on the shape of their average daily traffic patterns. It is found that using k-means clustering with five clusters yielded clearly separate types of traffic patterns which are then related to external spatial factors using a logistic regression model. A strong relationship between station type and spatial factors is found for all cities, and variations between the models for different cities are related to differences in commuting culture between cities. Average bike share demand for each station is modelled using a generalised linear model with a logarithmic link function, and coupled with the logistic regression model it is possible to predict average traffic patterns with reasonable precision.

Finally, in a case study of the Citi Bike system expansion in autumn 2019, the demand model is used to optimise station placement.