Identifying Usage Profiles of Station-Based Car-Sharing Members Using Cluster Analyses

With the growing usage of the internet, the possibility for shared mobility has risen just as much.
Beside ride-sharing, bike-sharing, and shared parking, this applies, especially to car-sharing.
Past research activities have often been limited to the economic, ecological, and urban
benefits of car-sharing, such as the number of privately owned cars that could be replaced by car-
sharing vehicles or the potential to save parking space. These analyses disregard the user’s
behavior and patterns of usage. However, to analyze, e.g., future market shares of car-sharing, we
first have to evaluate how car-sharing members use car-sharing and what purposes the trips might
serve. One such study has been conducted in Germany, however, using free-floating car-sharing
data.
Our focus of research is put on data from a station-based car-sharing provider and what
kind of user or usage profiles can be identified. We investigated this by performing a cluster
analysis using the k-means algorithm. The results indicate that there are five types of station-based
car-sharing users and usage respectively. There are commercial users, users who use car-sharing
... mehr
for regular and users who use it for irregular activities. Furthermore, car-sharing vehicles are used
to replace a second car and also for long distance travels.
These findings are in part consistent with the study on free-floating car-sharing but also
show some dissimilarities, as to be expected since the two systems generally serve different
purposes.