Measuring the vulnerability of bike-sharing system

Abstract

As a complex system, the bike-sharing system suffers from system failures, which can increase travel costs and impair user satisfaction. We proposed a concept of the vulnerability of bike-sharing system and a method to measure it. The method depends on the cost changes due to additional travel time induced by the failure of bike docking stations. It can capture the traffic mode transfer in the context of multi-modal traffic system, such as walking, bus, and subway. Moreover, to investigate the impact of network structure on the vulnerability, we developed the centrality measuring methods, and a community detection model for the bike-sharing system. Subsequently, the proposed methods are applied to Citi Bike in New York City, the largest bike-sharing system in the USA. The results show that the most vulnerable bike docking stations are located far from bus and railway stations, with low docking station density in their surrounding areas. We also found that the number of nearby bicycle stations, bus stops, and subway stations have a negative correlation with the vulnerability index. In contrast, the degree centrality and trip betweenness centrality are positively associated with the index. The proposed vulnerability analysis method can help urban planners to evaluate the design of a bike-sharing system and buttress operators to optimize maintenance planning. • A vulnerability measurement method for bike-sharing systems is proposed. • The method is formulated in the context of multi-modal traffic. • The centrality measuring and community detection models are built for impact studies. correlations. • The methods are applied to the bike-sharing system of Citi Bike. • Correlations of the key positive and negative factors with vulnerability are advised.