Data-driven Shuttle Service Optimization For Last-mile Transportation

Public transit rapidly blooms and leads to an increase in demand over the past decades,which places high stress on last/first-mile transportation.If the demand can not be met properly,travelers may drive to their destinations directly,which may result in traffic congestion and reducing a great amount of rail transit usage.Shuttle service has been considered as an efficient mobility option to cover the gap left by public transit,which is a critical extension to traditional public transit systems.To better design an efficient shuttle service system,determining the suitable service areas for allocating shuttle service resources based on demand features is the first and most complex step in the planning procedure.Considering the importance and challenges of planning shuttle service,the focus of this paper is on the determination of where to provide shuttle service and how to allocate the shuttle resource.To achieve this target,data-driven methods are applied to design the shuttle service for last/first-mile transportation system,which mainly comprises three parts.Firstly,an adaptive algorithm is given to identify the last/first-mile demand from urban data and then demand analysis is carried out to determine what impacts affect last/first-mile travel choice.Secondly,a data-driven framework is built to assess the performance of shuttle service in each area and determine the most suitable area to provide shuttle service.Finally,another data-driven framework is applied to simulate the process of responsive shuttle service and then optimize the fleet size with the objective of maximizing the operator profits.Using sharing-bicycles data,Shanghai is selected as the test area to verify the proposed framework,aiming to identify the most suitable metro-related area for providing shuttle service and then give an optimal allocation policy of shuttle service resource.The methods proposed in this paper provide a theoretical reference for the design of shuttle service,which gives an effective solution to promote the multimodal and sustainable public transportation system.