A route planning mechanism for supermarket shuttle service based on taxi traces. (March 2021)

• Record Type: Journal Article
• Authors: Yang, Guangfei; Yuan, Erbiao; Zhang, Xiang; Zhou, Huiyu
• Is Part Of: Research in transportation business & management. Volume 38(2021)
• Publication Date: 2021-03
• Subjects: Route planning -- Shuttle service -- Taxi GPS traces; Transportation -- Research -- Periodicals; Transportation -- Management -- Periodicals; Transportation -- Management; Transportation -- Research; Periodicals; 388.068
• DOI: 10.1016/j.rtbm.2020.100502
• Languages: English
• ISSNs: 2210-5395

Abstract

Abstract: With the rapid growth of retailing during the modernization of China, there is an increasing demand to improve the service quality of supermarkets. The supermarket shuttle service can have a direct impact on extending supermarket access, increasing shared transports, and improving customers' satisfaction. However, it is an open problem to probe the exact locations of potential customers and optimize the route for shuttle service. In this paper, we propose a route planning mechanism for supermarket shuttle service based on exploring the pick-up/drop-off information from large scale taxi GPS traces. Taxi GPS traces can inform us the human mobility patterns related to point of interests. Compared with traditional customer survey to collect information, the proposed approach can provide a more easy-to-deploy, efficient and effective solution. Based on the rich locational information of potential customers, we detect the most suitable bus stops of shuttle and optimize the routes to maximize the benefits for all stakeholders. The experimental results on real-life taxi data demonstrate the applicability as well as effectiveness of the proposed approach and exhibit promising performance of improve the quality of supermarket shuttle service. Highlights: The proposed DBSCAN-PAM algorithm generates feasible candidate bus stations more accurately. Use of genetic algorithm to select best routes with high efficiency and low cost. Optimization of genetic algorithm parameters to improve the accuracy of route generation calculations. Implications and suggestions to improve supermarket and government management.