Dynamic Redeployment to Counter Congestion or Starvation in Vehicle Sharing Systems

Abstract

Extensive usage of private vehicles has led to increased traffic congestion, carbon emissions, and usage of non-renewable resources. These concerns have led to the wide adoption of vehicle sharing (ex: bike sharing, car sharing) systems in many cities of the world. In vehicle-sharing systems, base stations (ex: docking stations for bikes) are strategically placed throughout a city and each of the base stations contain a pre-determined number of vehicles at the beginning of each day. Due to the stochastic and individualistic movement of customers,there is typically either congestion (more than required)or starvation (fewer than required) of vehicles at certain base stations. As demonstrated in our experimental results, this happens often and can cause a significant loss in demand. We propose to dynamically redeploy idle vehicles using carriers so as to minimize lost de-mand or alternatively maximize revenue for the vehicle sharing company. To that end, we contribute an optimization formulation to jointly address the redeploy-ment (of vehicles) and routing (of carriers) problemsand provide two approaches that rely on decomposability and abstraction of problem domains to reduce the computation time significantly. Finally, we demonstrate the utility of our approaches on two real world data sets of bike-sharing companies.

Online Repositioning in Bike Sharing Systems

“…3. Expected Sample Offline Policy Generation with revenue as objective (ESOF-Rev) (Ghosh et al 2015): As the objective of the formulation is to maximize the revenue, the approach tries to minimize the cost of vehicle movement in addition to minimizing lost demand. We compare the lost demand values and fuel cost with this approach.…”

“…In this section, we provide a generic model for representing the problem of repositioning and routing while considering demand uncertainty in bike sharing systems. We extend the expected demand based model of Dynamic Repositioning and Routing Problem (DRRP) provided by Ghosh et al (2015). Here, we provide the Dynamic Repositioning and Routing Problem with Demand Uncertainty (DR-RPDU), that considers multiple samples of demand at different stations and timesteps.…”

“…Decomposition across samples does not help in reducing computation time, as for problems with large number of stations, even for a single sample MSS takes a long time (thousands of seconds). Therefore, we extend the Lagrangian Dual Decomposition (LDD) (Fisher 1985) method proposed by Ghosh et al (2015) for solving offline repositioning and routing problem in bike sharing. Our key contributions within this LDD approach when applied to DRRPDU are two fold:(i) We update LDD to account for multiple demand samples; (ii) We significantly improve the computational complexity of solving the routing problem by Routing:…”

“…But they do not consider stations getting imbalanced during the day. Hence there is a need to focus on dynamic repositioning of bikes (Contardo, Morency, and Rousseau 2012;Ghosh et al 2015; during the day. Shu et al (2013) provide an optimization model for dynamic repositioning of bikes, but they do not consider routing of carrier vehicles.…”

“…Shu et al (2013) provide an optimization model for dynamic repositioning of bikes, but they do not consider routing of carrier vehicles. Ghosh et al (2015; consider the problem of dynamic repositioning of bikes along with the problem of finding the routing policy for carrier vehicles. They provide an offline policy generation approach based on mean demand computed from the historical data.…”