GIS-based Optimization of Bike-sharing Systems

GIS-based Optimization of Bike-sharing Systems


Matthias Benedek, Department of Geography, University of Augsburg, Carolin Von Groote-Bidlingmaier, Department of Geography, University of Augsburg, Sabine Timpf, Department of Geography, University of Augsburg


Bike-sharing systems (BSS) offer the possibility to avoid restrictions of public and individual transport. To increase the use of the BSS, an optimization is performed using a Geographic Information System (GIS).


Bike-sharing systems (BSS) expand the offer of mobility in many cities. For the target groups (citizens without bicycles, commuters or tourists) BSS offer the possibility to avoid the restrictions of public transport such as inflexible routes and departure times, and those of individual transport such as searching for parking spaces, pollution, and noise. BSS thus allow for choosing a flexible transport means for the occasion. There are two essential issues in order to provide potential users with a suitable offer. First, stations have to be optimally allocated according to need and second, the transport of bicycles due to daily fluctuations in demand should be quick and efficient.
An optimization strategy for the BSS is needed with the goal to always provide an adequate number of bicycles to the users, despite varying loan periods, single way use and strong fluctuations in demand that depend on the time of day and the day of the week. If the existing distribution of the stations does not allow for this self-regulatory “natural” flow of bicycles, the secondary goal is to organize a manual transport of bicycles with minimal costs. Furthermore, the positions of all stations should be studied in order to find inefficient stations as well as places where new stations might be needed.
The optimization process we propose is subdivided into two main steps. The first step initially defines appropriate dimensions of analysis, such as usage data, population density and points of interest (POI), for the purpose of evaluating stations in terms of efficiency. Subsequently, a catchment area analysis is used to identify inefficient positioning of stations. Finally, new or alternative locations for stations are searched and compared to the initial situation using network analysis within a Geographic Information System.
The second step involves the logistics regarding the availability of bicycles. The aim here is to minimize the costs for the operator and to provide a better offer to the users at the same time. Therefore, usage and demand of the stations is analyzed and a route (as required: fastest or shortest path) computed to transport the bicycles to locations, where they will be needed. Stations along this route that were initially not part of the reallocation plan will be included in the resupply.


Association for European Transport