Intelligent order matching for uber-like shareable vehicle systems

Abstract

We present a novel order dispatch algorithm in large-scale on-demand ride-hailing platforms that take account of the dynamic characteristics associated with workers. Although most traditional order dispatch approaches generally focus on providing a better user experience for passengers and maximizing revenue by optimizing resource utilization, the proposed algorithm is designed to take into an account of the collective productivity of all workers and maximizing it opportunistically in response to stochastic changes in situational factors. This is also accompanied by a Multi-Agent Simulation to simulate the complex action and interactions of the drivers and passengers, and to analyze the effects of change in factors. After the implementation of the algorithm and simulation, we evaluated the effects in earnings, reputation and fatigue. In the most recent outbreak of the disease on COVID-19, the simulation also has a few mechanisms in showing how it spread among the drivers and passengers through the use of the proposed algorithm.