Carpooling Service for Large-Scale Taxicab Networks.
Desheng Zhang,Tian He,Fan Zhang,Mingming Lu,Yunhuai Liu,Haengju Lee,Sang H. Son
DOI: https://doi.org/10.1145/2897517
2016-01-01
ACM Transactions on Sensor Networks
Abstract:Carpooling has long held the promise of reducing gas consumption by decreasing mileage to deliver coriders. Although ad hoc carpools already exist in the real world through private arrangements, little research on the topic has been done. In this article, we present the first systematic work to design, implement, and evaluate a carpool service, called coRide, in a large-scale taxicab network intended to reduce total mileage for less gas consumption. Our coRide system consists of three components, a dispatching cloud server, passenger clients, and an onboard customized device, called TaxiBox. In the coRide design, in response to the delivery requests of passengers, dispatching cloud servers calculate cost-efficient carpool routes for taxicab drivers and thus lower fares for the individual passengers. To improve coRide’s efficiency in mileage reduction, we formulate an NP-hard route calculation problem under different practical constraints. We then provide (1) an optimal algorithm using Linear Programming, (2) a 2-approximation algorithm with a polynomial complexity, and (3) its corresponding online version with a linear complexity. To encourage coRide’s adoption, we present a win-win fare model as the incentive mechanism for passengers and drivers to participate. We test the performance of coRide by a comprehensive evaluation with a real-world trial implementation and a data-driven simulation with 14,000 taxi data from the Chinese city Shenzhen. The results show that compared with the ground truth, our service can reduce 33% of total mileage; with our win-win fare model, we can lower passenger fares by 49% and simultaneously increase driver profit by 76%.