Minimizing fleet size and improving vehicle allocation of shared mobility under future uncertainty: A case study of bike sharing.

As a rapidly expanding type of shared mobility, bike sharing is facing severe challenges of bike over-supply and demand fluctuation in many Chinese cities. In this paper, a large-scale method is developed to determine the minimum fleet size under future demand uncertainty, which is applied in a case study with millions of bike sharing trips in Nanjing. The findings show that if future uncertainty is not considered, more than 12% of trip demands may not be satisfied. Nevertheless, the proposed algorithm for minimizing fleet size based on historical trip data is effective in handling future uncertainty. For a bike sharing system, supplying 14.5% of the original fleet could be sufficient to meet 96.8% of trip demands. Meanwhile, the results suggest a unified platform that integrates multiple companies can significantly reduce the total fleet size by 44.6%. Moreover, in view of the Coronavirus Disease 2019 (COVID-19) pandemic, this paper proposes a contact delay policy that maintains a suitable usage interval, which results in increased bike amount requirements. These findings provide useful insights for improving resource efficiency and operational services in shared mobility applications. [Display omitted] • Our method can minimize the fleet size for millions of trips in large size networks. • Our improved algorithm can handle the uncertainty of future trip demands. • Supplying 14.5% of the original fleet can meet 96.8% of dockless bike sharing trips. • A contact delay policy is proposed and tested for the COVID-19 pandemic response. • Integrating multiple companies into a unified platform can reduce fleet size. [ABSTRACT FROM AUTHOR]