Abstract
This article presents an analysis of the supply and demand of a FFBSS recently implemented in the city of Bologna, Italy. The main aspects treated in this paper are: analysis of bike availability; temporal analysis of FFBSS demand; calibration and validation of a novel model that predicts the number of daily trips per available bike. This model is based on a linear combination of several day attributes, including meteorological and day-type attributes. Moreover, an origin to destination analysis is generated showing the spatial distribution of FFBSS trips. The methods are applied to a scenario with almost a million GPS traces recorded between July and October 2018 by the FFBSS in Bologna. Findings could support FFBSS companies to better understand the fluctuation of both the transport demand and supply of this relatively recent transport mode, as to make more efficient decisions when distributing or relocating bicycles.
Highlights
Congestion problems in urban centers, together with the issue of environmental sustainability, has led the European Union to develop regulations aimed at development of efficient and sustainable urban mobility system
An analysis of the first four months - from 1st July to 31th October - of the first and only FFBSS activity in Bologna has been performed in order to characterize and interpret the FFBSS supply and usage
Since the FFBSS demand is not yet widely studied in literature, this research sheds some light on the supply and demand side of FFBSS, which may help in the planning and operation of such systems
Summary
Congestion problems in urban centers, together with the issue of environmental sustainability, has led the European Union to develop regulations aimed at development of efficient and sustainable urban mobility system. The main issue for freefloating bike-sharing system is that reallocation models require the demand-pattern in order to predict where to reposition bikes [10,11] - an issue not yet widely studied in literature. A poorly designed FFBS system, e.g. in terms of recommended parking lots, or without providing discount for the correct bikeusage, could lead to higher travel-costs for the user and discourage the use of the service [23] These problems of the bike-sharing system are not yet widely studied in literature and to date no FFBSS demand model has been studied in detail. The model would allow to predict the FFBSS demand for the following day according to the day-type attributes and weather conditions Such predictions are useful for the scheduling of the bike re-allocation.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
