An increasing number of large cities worldwide are adopting bicycle-sharing schemes as a means to reduce vehicular traffic, pollution, and energy consumption. Several papers have studied vehicle-sharing systems in order to consider economic aspects, growth trends, or environmental impact. The research challenges here are the construction of efficient redistribution policies for vehicles and the creation of incentive mechanisms to improve load balancing within these systems.
Customer satisfaction plays an important role. Operators of bicycle-sharing schemes want to avoid disappointing experiences such as a user wanting to hire a bicycle finding a station empty, or a user with a bicycle finding a station full when wanting to return their bicycle. In both cases, this involves moving to a nearby station and adding delays to the journey.
Up to now, only a very few simple analytical models have been developed to approach the problem of service provisioning for bicycle-sharing schemes. In particular, these models often ignore spatial aspects or daily patterns inherent to these systems. In the QUANTICOL project our approach will be complementary to the aforementioned ones in that it may be used during early stages of the design. Our prediction model will make it possible to tackle simultaneously both the long-term problem of capacity planning (e.g. the size and placement of stations) and short-term operational decisions like incentives mechanisms or efficient vehicle repositioning strategies.
We will present details of modelling studies of bike sharing systems to investigate the impact of structural changes such as removal or addition of bike repositories, or the capacity within repositories. We will use these studies to support the construction of efficient redistribution policies and incentive mechanisms to improve load balancing.