Exploring Patterns in the Free-Floating Bike Sharing Usage by Analysing GPS Traces

The recent evolution of free-floating bike-sharing system (FFBSS) is a result of the ever growing need to introduce sustainable urban mobility solutions. However, the factors that influence the supply and demand of such systems are not yet widely studied in literature. The paper aims at expanding the knowledge framework of the FFBSSs, by thoroughly analyzing a big data set composed of almost a million GPS traces, which were recorded between July and October 2018 from the MOBIKE FFBSS in the city of Bologna, Italy. The methodology consists in a detailed analysis of the available data set, designed to quantify the FFBSS demand and supply, while focusing on the FFBSS demand side. The GPS traces have been imported, filtered and analyzed with SUMOPy, a GIS software developed at the University of Bologna. Special software-extensions for the present analysis have also been implemented in Python. Regarding the analyses of the supply side, it became apparent that one of the worst problems of the FFBSS in Bologna may be vandalism: the number of shared bikes in circulation decreased exponentially in time. On the demand side, a model has been calibrated that predicts the daily FFBSS demand as a function of several attributes, of which the most significant have been: weather conditions (rainy day or not), daily average temperature, day-type (weekday or holiday) and period (working or holiday). However, unfavourable weather conditions seem to decrease only temporarily the demand; in fact, after such events, the bike activity restarts immediately at full capacity. A spatial analysis confirms the quite obvious findings from previous studies indicating that trips are more frequents in the city centre and decrease towards the suburb. Moreover, the evolution of the bike-usage during the day has been analyzed for different types of traffic zones and the results show that these latter could be clustered in different classes regarding a series of characteristics, such as the frequency of use and the tendency towards either accumulate or disperse bicycles. Note that this study has been limited to analyse only the satisfied demand, which is constrained by the limited bike supply. Nevertheless, the number of estimated daily trips compared with the effective trip numbers have shown a high R-square, proving the validity of this analysis. These findings could support FFBSS companies to better understand the fluctuation of both, the transport demand and supply of this recent transport system, in order to improve the distribution and relocation of their bikes.

  • Record URL:
  • Supplemental Notes:
    • Abstract used by permission of Association for European Transport. Alternative title: Prediction of the Free Floating Bike Sharing System Demand from GPS traces
  • Corporate Authors:

    Association for European Transport (AET)

    1 Vernon Mews, Vernon Street, West Kensington
    London W14 0RL,    
  • Authors:
    • Poliziani, Cristian
    • Rupi, Federico
    • Schweizer, Joerg
  • Conference:
  • Publication Date: 2020


  • English

Media Info

  • Media Type: Digital/other
  • Features: Figures; References; Tables;
  • Pagination: 18p
  • Monograph Title: European Transport Conference 2020

Subject/Index Terms

Filing Info

  • Accession Number: 01768529
  • Record Type: Publication
  • Files: TRIS
  • Created Date: Feb 17 2021 1:55PM